Managing the "Republic of NGOs": Accountability and Legitimation Problems Facing the UN Cluster System

Vanderbilt Journal of Transnational Law Vanderbilt Journal of Transnational Law

Volume 47 Issue 1 January 2014

Managing the "Republic of NGOs": Accountability and Managing the "Republic of NGOs": Accountability and

Legitimation Problems Facing the UN Cluster System Legitimation Problems Facing the UN Cluster System

J. Benton Heath

Article 4

Part of the International Humanitarian Law Commons

Managing the "Republic of NGOs": Accountability and Legitimation Problems Facing the UN Cluster System

J. Benton Heath*

ABSTRACT

This Article critically assesses the crucial but troubled system for the coordination of international humanitarian assistance-the UN "cluster approach." Regardless of whether the cluster approach actually helps in disaster response, it exercises substantial power over affected populations by assigning competences and leadership roles. The built-in mechanisms for controlling this power are unworkable because they ultimately fail to resolve the tension between humanitarian organizations' autonomy and the need for coordination. This Article identifies the emergence of an alternative model of accountability, based on mutual monitoring and peer review. Drawing on governance and experimentalism, this Article teases out the institutional and normative implications of such a model. In particular, this Article argues, the emerging turn toward peer review in the cluster approach would demand dramatic improvements to the system's inclusion of affected populations. This investigation may carry broader lessons for transnational networks and the study of accountability in global governance.

theories of

network

The author is presently a judicial clerk for Judge Robert D. Sack of the U.S. Court of Appeals, Second Circuit, in New York, and holds a J.D. and LL.M. from New York University School of Law. Many thanks to Philip Alston, Julian Arato, Grdinne de Birca, Lorenzo Casini, Michael Cooper, Georgios Dimitropoulos, Adrian DiGiovanni, Jasper Finke, Kelly Geoghegan, Carla M. Greenberg, Ellie Happel, Benedict Kingsbury, Hari Osofsky, Dirk Salomons, Margaret Satterthwaite, Richard B. Stewart, Abby Stoddard, Ren6 Uruefia, and Eduardo Valenica-Ospina for insightful comments and guidance. Thanks also to the staff of the Vanderbilt Journal of Transnational Law for thorough and conscientious editing. All remaining errors are the author's responsibility. Please note that this Article is written in the author's personal capacity and does not reflect the opinions of his employer.

VANDERBILTIOURNAL OF TRANSNATIONAL LAW

TABLE OF CONTENTS

I II.

III.

IV.

V.

VI.

...............................

INTRODUCTION THE RISE OF THE UNITED NATIONS AS A HUMANITARIAN COORDINATOR THE POWER OF HUMANITARIAN COORDINATION: OPERATIONAL AND INSTITUTIONAL ASPECTS ............. Trampling Diversity? Harmonization of A. Policy in the Clusters Problematizing Diversity: Variance Among Leaders

.....................

......................

.......................

...........

ACCOUNTABILITY, LEGITIMACY, AND INSTITUTIONAL DESIGN . A.

...................................

A Broad Approach to the Accountability of Networks.............................. Accountability as the Self-Justification of Institutions

............................

.........

.......

........................

SUPERVISORY ACCOUNTABILITY IN THE UN HUMANITARIAN ARCHITECTURE A.

Autonomy and Coordination as Constraints on Accountability Outlines of a Formal Supervisory Structure..... The Breakdown of the Formal Structure...........

B. C. THE PROMISES AND CHALLENGES OF PEER REVIEW IN THE CLUSTER SYSTEM A.

..............................

..................

Peer Review and the Autonomy-Coordination Tension................................ Experimental Humanitarian Institutions: Sketching an Alternative Model...... 1. Active and Clear Rulemaking 2. Overcoming the Peer Monitoring Deficit.... 3. Reflexivity of Policy Making..... ......... 4. Rethinking Hierarchy

.......................

........

................

C. Participation as the Central Problem of

a Reimagined System ....................... CONCLUSION......................................

VII.

268 271 273

281 284 286 286 288

289 292

MANAGING THE 'REPUBLIC OF NGOS"

I. INTRODUCTION

The demand for improved coordination marks all contemporary discussions of humanitarian aid.' Three years after the devastating earthquake in Haiti, coordination failures take the blame for the interminable pace of the recovery effort.2 Meanwhile, one observer has defined the "main problem" of aid delivery in rebel-held Syria as "the absence of coordination" between various local and foreign actors.3 The lack of coordination can lead to wasted resources, needless self-described humanitarian organizations continue to proliferate, and as the complexity of

suffering, or both.4 As

See, e.g., MICHAEL BARNETT, EMPIRE OF HUMANITY: A HISTORY OF HUMANITARIANISM 235 (2011) (noting with skepticism the trend toward calls for improved coordination); SUE GRAVES, VICTORIA WHEELER & ELLEN MARTIN, LOST IN TRANSLATION: MANAGING COORDINATION AND LEADERSHIP REFORM THE HUMANITARIAN SYSTEM 1 (Humanitarian Pol'y Grp., Pol'y Brief No. 27, 2007) ("Leadership and coordination are integral to maximising the impact of multiple aid providers, but decades of effort in this area have produced mixed results at the field level."); BRUCE D. JONES, THE CHANGING ROLE OF THE U.N. IN PROTRACTED CRISES 3 (Humanitarian Pol'y Grp., Research Briefing No. 17, 2004) (noting a growing belief that coordination problems were to blame for failures in protracted crises); Laura M. E. Sheridan, Note, Institutional Arrangements for the Coordination of Humanitarian Assistance in Complex Emergencies of Forced Migration, 14 GEO. IMMIG. L.J. 941, 94344 (2000) (examining institutional developments in this area throughout the latter half of the twentieth century).

2.

See, e.g., Deborah Sontag, Rebuilding Lags in Haiti After Billions in PostQuake Aid, N.Y. TIMES, Dec. 24, 2012, at Al, available at http://www.nytimes.com/ 2012/12/24/worldlamericas/in-aiding-quake-battered-haiti-lofty-hopes-and-hard-truths. html?pagewanted=all rebuilding); Neil MacFarquhar, U.N. Is Faulted as Lacking Coordination of Aid and Security in Haiti, N.Y. TIMES, March 3, 2010, at All; David Bressan, Post-Disaster Recovery: Lessons from the 2010 Haiti Earthquake, SCIENTIFIC AMERICAN HISTORY OF GEOLOGY BLOG (Jan. 12, 2012), recovery-lessons-from-the-2010-haiti-earthquake/ (stressing coordination problems in the allocation of funding for reconstruction projects).

http://blogs.scientificamerican.com/history-of-geology/2012/01/12/post-disaster-

(noting an unsystematic

approach

3.

Kristen McTighe, Syrian Insurgents Say Aid Isn't Getting Where It Needs to Go, INT'L HERALD TRIBUNE (Mar. 7, 2013), http://www.nytimes.com/2013/03/07/world/ middleeast/syrian-insurgents-say-aid-isnt-getting-where-it-needs-to-go.html.

4.

To give just one example, a 2007 Red Cross and Red Crescent study recounts the story of an unidentified NGO that vaccinated some children in an Indonesian village following the Indian Ocean Tsunami, "leaving no records and no way to determine who had been vaccinated and who had not." DAVID FISHER, INT'L FED'N OF RED CROSS & RED CRESCENT SOCIETIES, LAW AND LEGAL ISSUES IN INTERNATIONAL DISASTER RESPONSE: A DESK STUDY 14 (2007). This failure to leave a record of one's activities forces the next aid agency to either waste resources on individuals who were likely already vaccinated or treat the entire population as vaccinated, accepting the risk that some may not be. To understand the possible effects of coordination failures, one need only imagine endless variations of this scenario, with virtually no limit as to scale or complexity.

[VOL. 47239 [

disaster response emerges as a growing concern, for sustained and reliable coordination mechanisms is both pressing and uncontestable.5

the need

Institutions are emerging to meet this demand, and there is no shortage of coordination mechanisms in Haiti or any other disasterravaged nation. Chief among these mechanisms is the UN system for coordinating humanitarian operations, generally known as the "cluster approach."6 Though no more than a loose network connecting autonomous organizations, the cluster approach may be understood as an institution in its own right, one that exercises substantial power in disaster-affected states.7 Whatever its flaws, a coordination apparatus like the cluster system is the connective tissue that brings together various humanitarian actors, forming what some Haitians have derisively called the "Republic of NGOs."8

This Article critically assesses the power, accountability, and legitimation of a coordination network such as the cluster approach.9 In so doing, it brackets the question of whether this system actually improves coordination, or how improvement would be measured.10 This Article explains that the cluster approach alters the realities of disaster response by assigning competences and leadership roles, setting policies, and channeling funding. Embedded within arguments regarding the accountability of this power structure are deeper questions concerning the legitimation of an identifying and institution such as

the cluster approach. By

such

5. 28-31. 6.

On the proliferation of international humanitarian organizations, see id. at

See generally What Is the Cluster Approach?, HUMANITARIANRESPONSE,

https://clusters.humanitarianresponse.info/about-clusters/what-is-the-cluster-approach (last visited Dec. 24, 2013).

7.

On the importance of studying networks as actors in their own right, see generally MARGARET E. KECK & KATHRYN SIKKINK, ACTIVISTS BEYOND BORDERS: ADVOCACY NETWORKS IN INTERNATIONAL POLITICS (1998). The accountability of such networks emerges as a central concern in Anne-Marie Slaughter, Accountability of Global Governance Networks, 8 IND. J. GLOBAL LEGAL STUD. 347 (2001).

8.

E.g., Madeline Kristoff & Liz Paranelli, Haiti: A Republic of NGOs?, UNITED STATES INSTITUTE FOR PEACE: PEACEBRIEF, Apr. 26, 2010, at 1, available at http://www.usip.org/files/resources/PB%/ 2023%20Haiti%20a%2Republic%20of%2ONG Os.pdf; Marjorie Valbrun, After the Quake, Praise Becomes Resentment in Haiti: Amid a Slow Recovery, Haitians Question the Work of Aid Groups, CTR. FOR PUB. INTEGRITY (last updated Jan. 12, 2012, 5:44 PM), http://www.publicintegrity.org/ 2012/01/10/7838/afterquake-praise-becomes-resentment-haiti.

9.

Should one require a definition of power, a broad definition, which refers generally to the human capacity to transform states of affairs, may be appropriate. See, e.g., ANTHONY GIDDENS, NEW RULES OF SOCIOLOGICAL METHOD: A POSITIVE CRITIQUE OF INTERPRETATIVE SOCIOLOGIES 111 (1976) (defining power as the "transformative capacity of human agency").

10.

Five years after the implementation of the system, an external evaluation found that "it is hard to pin down exactly how the cluster approach was or is intended to work." JULIA STEE'i'S ET AL., INTER-AGENCY STANDING COMM., CLUSTER APPROACH EVALUATION 2: SYNTHESIS REPORT 24 (2010).

interpreting emerging trends in this discourse, this Article draws out the normative choices that face the UN humanitarian architecture.

This Article argues that changes within the cluster approach are placing increasing stress on the inclusion of voices from vulnerable and disaster-affected communities. Voice is not a new concern for humanitarian actors." But institutional developments within the cluster system, designed to ensure its accountability and legitimacy, raise the problem of voice and inclusion to a place of primacy, without necessarily solving it. As humanitarian actors begin to unlock the potential of the cluster system for decentralized learning and experimentation, the fate of that system will be increasingly bound to the question of how well it secures the effective participation of local populations.

Despite a renewed interest in disaster response law,'2 the institutions through which disaster response is operationalized remain understudied in the legal literature.13 This deficit is at least partly attributable to the orthodox view of public international law in this area, which focused on rules of conduct and responsibility, rather than institutional structure.14 But the intensive

remains

11.

See, e.g., Georg Frerks & Dorothea Hilhorst, Evaluation of Humanitarian Assistance in Emergency Situations 15-16 (UNHCR Evaluation & Pol'y Analysis Unit, (detailing persisting challenges associated with Working Paper No. 56, 2002) participatory approaches to aid delivery); Amanda M. Klasing, P. Scott Moses & Margaret Satterthwaite, Measuring the Way Forward in Haiti: Grounding Disaster Relief in the Legal Framework of Human Rights, 13 HEALTH & HUM. RTS. 1, 13-16 (2011) (discussing diverging attitudes toward participation among aid workers).

12.

These have included recent conferences under the auspices of the "Four Societies" of international law, the Canadian Council on International Law, and the Hague Academy, as well as ongoing codification efforts by the International Federation of Red Cross and Red Crescent Societies and the International Law Commission.

13.

See Nadia Khoury, Hague Academy Examines the International Law of Catastrophe, INT'L FED'N OF RED CROSS & RED CRESCENT SOCIETIES (Feb. 27, 2012, 1:25 PM), https://www.ifrc.org/en/what-we-do/idrl/latest-news/disaster-law-newsletterfebruary-2012/hague-academy-of-international-law-seminar-on-the-legal-challenges-ofcatastrophes/ (interview with Professor Samantha Besson, who points to the need for further analysis of institutions in disaster response). But see Giovanni de Siervo, Actors, Activities and Coordination in Emergencies, in INTERNATIONAL DISASTER RESPONSE LAW 485 (Andrea de Guttry, Marco Gestri & Gabriella Venturini eds., 2012) (addressing relevant actors in humanitarian practice). Other recent works have begun to tackle areas of disaster response that require some sensitivity to an institutional perspective. See, e.g., David Fisher, Legal Implementation of Human Rights Obligations to Prevent Displacement Due to Natural Disasters, 41 STUD. TRANSNAT'L LEGAL POL'Y 551, 561-64 (2010) (addressing what might be categorized as compliance issues); Jim Chen, Modern Disaster Theory: Evaluating Disaster Law As a Portfolio of Legal Rules, 25 EMORY INT'L L. REV. 1121, 1131-42 (2011) (applying a riskmanagement theory to the assessment of disaster rules).

14.

Moreover, the leading treatises on international institutional law spend little time addressing operational activities or emergency situations. It has recently been suggested that emergency powers of international organizations should become a subfield of study in the field of global administrative law, though few works have followed in that vein. Benedict Kingsbury & Lorenzo Casini, Global Administrative

in particular

study of institutional design, and the normative assessment of institutional structure, is a fundamental task in the history of public law.15 The task of the public lawyer should be to draw out the normative implications of a particular practice through the careful and context-sensitive study of institutional structure and operations, not to develop policy prescriptions through the direct application of abstract theories.16 In undertaking this approach, this Article aims to make a small contribution to the effort to reclaim the study of institutions for legal scholars.

The particular institution studied here is a relatively new invention. Though coordination is a persistent historical problem of humanitarianism,17 the United Nations has emerged as a central player two decades.'8 In 2005, following the highly complex response to the Indian Ocean tsunami and the crisis in Darfur, a comprehensive review of the humanitarian system recommended the designation of

in humanitarian coordination

the past

largely

Law Dimensions of International Organizations Law, 6 INT'L ORG. L. REV. 319, 334-38 (2009); see also Joost Pauwelyn & Ayelet Berman, Emergency Action by the WTO Director-General: Global Administrative Law and the WTO's Initial Response to the 2008-09 Financial Crisis, 6 INT'L ORG. L. REV. 499, 506-08 (2009) (exploring arguments that the WTO exceeded its mandate in response to the financial crisis); Ulrich Garms, Promoting Human Rights in the Administration of Justice in Southern Sudan: Mandate and Accountability Dilemmas in the Field Work of a DPKO Human Rights Officer, 6 IN'I"L ORG. L. REV. 581, 594-600 (2009) (assessing the limitations on accountability in UN peacekeeping operations).

15.

For one comprehensive effort to take a public-law approach to international institutions, see generally THE EXERCISE OF PUBLIC AUTHORITY BY INTERNATIONAL INSTITUTIONS: ADVANCING INTERNATIONAL INSTITUTIONAL LAW (Armin von Bogdandy et al. eds., 2010).

16.

One recent work describes such a method as:

[A] more faithful to and respectful of law than many of the alternatives. Rather than philosophy treating law as merely a tool with which to implement the conclusions of an extralegal philosophical inquiry-dropping like an imperious and alien visitor, delivering pronouncements, and flying off againthe social-theoretic approach takes law as a form of embodied ethical life with a certain immanent moral content already in place, which philosophy can help bring to light and expose to question.

Joshua Kleinfeld, A Theory of Criminal Victimization, 65 STAN. L. REV. 1087, 1152 (2013). 17.

It may be among the oldest and most intractable problems of modern humanitarianism. See Convention and Statute Establishing an International Relief Union art. 2(2), July 12, 1927, 1932 L.N.T.S. 249 (no longer in force) (stating that the Union's purpose will be "to co-ordinate ... the efforts made by relief organizations"); Twenty-First Int'l Conference of the Red Cross, Istanbul, Turk., Sept 1969, Declaration of Principles for International Humanitarian Relief to the Civilian Population in Disaster Situations, Res. XXVI, T 3 (committing to "secure prompt action and effective allocation of resources and to avoid duplication of effort").

18.

The contemporary era began following the Kurdish refugee crisis caused by the first Gulf War. Tom J. Farer & Felice Gaer, The UN and Human Rights: At the End of the Beginning, in UNITED NATIONS, DIVIDED WORLD 240, 255-57 (Adam Roberts & Benedict Kingsbury eds., 2d ed. 1993).

lead agencies in areas where coordination was weak.19 The next year, the Inter-Agency Standing Committee (IASC) finalized guidelines for such an approach, designating nine "cluster leads" in substantive areas where "there are clearly identified gaps in capacity."20 The socalled cluster approach now claims to be the central method for coordinating humanitarian activity experiencing overwhelming disasters. This system is the focus of this Article, though any future arrangement will have to confront problems similar to those described here, unless the realities of humanitarian institutions are dramatically altered.

in countries

This Article does four things. Following an introduction to the humanitarian system (Part II), it identifies a form of power defined as "institutional choice" and argues that the mere fact of assigning competences and tasks to one organization rather than another has practical consequences at all levels of action (Part III). Second, it relates the discourse of accountability in the cluster approach to the difficulties in controlling and correcting the effects of institutional choice. In the process, the Article sketches an approach to institutions that melds normative and conceptual inquiry with a detailed, practical focus on the actual design and workings of institutions (Part IV). Third, it applies that method to critique the vertical, hierarchical accountability structure that was designed for the cluster approach, demonstrating that the system's operation in practice has forced a confrontation with its normative shortcomings (Part V).

Fourth, this Article accepts the invitation, posed by some recent reports from practitioners and consultants, to reimagine the system for humanitarian coordination in terms of a horizontal accountability structure (Part VI). Several observers of humanitarian practice have suggested that processes to encourage mutual monitoring, continuous peer review, and experimentation might remedy the deficit in effective oversight and control.2' Drawing on insights from recent literature on network governance and democratic experimentalism,22 this Article teases out the institutional and normative implications of such an approach. Though this approach is promising, this Article argues that a horizontal structure represents a high-risk strategy for

19.

HUMANITARIAN RESPONSE REVIEW 51 (commissioned by United Nations Emergency Relief Coordinator and Under-Secretary-General for Humanitarian Affairs, Aug. 2005) [hereinafter HUMANITARIAN RESPONSE REVIEW].

20.

INTER-AGENCY STANDING COMM. [IASC], GUIDANCE NOTE ON USING THE CLUSTER APPROACH TO STRENGTHEN HUMANITARIAN RESPONSE 3 (Nov. 24, 2006) [hereinafter IASC, CLUSTER APPROACH]. "Substantive areas," refers to operational sectors, such as food, shelter, camp management, and refugee protection.

21. 22.

See infra text accompanying notes 182-90. For overviews of these concepts, see generally Charles F. Sabel & William H. Simon, Minimalism and Experimentalism in the Administrative State, 100 GEo. L.J. 53 (2011); Grdinne de Bdrca, New Governance and Experimentalism: An Introduction, 2010 WiS. L. REV. 227.

/VOL. 47-239

institution's humanitarian coordination, legitimacy ever more tightly to its ability to include and respond to the voices of affected populations.

one that hitches

II. THE RISE OF THE UNITED NATIONS AS A HUMANITARIAN

COORDINATOR

system

The cluster approach-the UN

for coordinating humanitarian activities-generates a framework loaded with the potential for the exercise of power, the expansion of organizational competences, and the generation of winners and losers. The reality is often more muted, as the power of the cluster system depends on its acceptance by relevant actors.23 Moreover, the relative importance of the UN framework will likely increase in inverse proportion to the involvement of the state government, which retains the primary responsibility for coordinating relief on its territory.24 This Part will explain the evolution of the cluster approach in order to tease out the various ways in which this structure exercises emergency power. The following Part then turns to a more in-depth analysis of the ways in which the structure facilitates or blocks the exercise of power by participating institutions.

The Office for the Coordination of Humanitarian Affairs (OCHA), an office of the UN Secretariat, manages the coordination of humanitarian affairs, overseeing humanitarian operations as well as the coordination of humanitarian policy and advocacy.25 Policy setting and best practices are developed by the IASC, which includes all the UN operational agencies, such as the High Commissioner for

23.

See, e.g., Miriam Stumpenhorst, Rolf Stumpenhorst & Oliver Razum, The UN OCHA Cluster Approach: Gaps Between Theory and Practice, 19 J. PUB. HEALTH 587, 588 (2011) (discussing the shortcomings of the cluster approach during the Haitian earthquake relief efforts due to the conflicting goals of independent relief organizations).

24.

On the primary role of the host state, see Report of the Int'l Law Comm'n, 62d Sess., Third Report on the Protection of Persons in the Event of Disasters, TT 7989, U.N. Doc. A/CN.4/629 (Mar. 31, 2010) (Edurardo Valencia-Ospina) (surveying treaty provisions to this effect); J. Benton Heath, Note, Disasters, Relief, and Neglect: The Duty to Accept Humanitarian Assistance and the Work of the International Law Commission, 43 N.Y.U. J. INVL L. & POL. 419, 453-54 (2011) (arguing for a rule that balances the state's human rights obligations to provide aid with its primary role).

25.

E.g., Valerie Amos, UNOCHA Organizational Diagram, UNITED NATIONS OFFICE FOR THE COORDINATION OF HUMANITARIAN AFFAIRS, http://ochanet.unocha.org/ plDocuments/UNOCHA%200rganigramme%202011.pdf (last visited Dec. 24, 2013). OCHA was first established (albeit under a different name) in 1991 by the secretarygeneral. Strengthening the Coordination of Humanitarian Emergency Assistance of the United Nations, G.A. Res. 46/182, Annex, TI1 34, 36, U.N. Doc. A/RES/46/182 (Dec. 19, 1991). OCHA received its current designation in 1998, after a reorganization that also expanded the office's mandate to include broader coordination functions. History of OCHA, UNITED NATIONS OFFICE FOR THE COORDINATION OF HUMANITARIAN AFFAIRS, http://www.unocha.orglabout-us/who-we-are/history (last visited Dec. 24, 2013).

MANAGING THE "REPUBLIC OF NGOS

2 247

Refugees (UNHCR) and the World Health Organization (WHO), and which extends standing invitations to a range of nongovernmental the International organizations (IOM).26 Country-level efforts are Organization overseen by an in-country humanitarian coordinator, who is named by the head of OCHA and remains directly accountable to her.27 The head of OCHA is an appointee of the UN secretary-general.28

(NGOs), for Migration

the Red Cross, and

in 2005

the previous

The cluster approach was developed

following complicated responses to the Indian Ocean tsunami and the crisis in Darfur.29 This new method of disaster management was meant to institutions.30 In solve several problems endemic particular, "clear operational in key sectors."3' The reforms accountability and introduced in 2005 responded to these complaints by introducing areas where clearly responsibility had previously been murky.32 The key innovation was the assignment of a lead agency for each sector, which would be "responsible for mapping needs, planning, monitoring, coordination and reporting," and would act as a "provider of last resort."33

to existing lacked

responsibilities

leadership

leadership

structure

defined

The cluster approach works at two levels: country level and global.34 At the global level, the clusters are standing bodies, where lead agencies coordinate standard setting, dissemination of best practices, and capacity building among responders.35 Country-level

26.

See Strengthening

the Coordination of Humanitarian Emergency 38, U.N. Doc. Assistance of the United Nations, G.A. Res. 46/182, Annex, A/RES/46/182 (Dec. 19, 1991) (establishing the IASC). Current members of the IASC, which is crucial for setting policy on humanitarian coordination, are the Food and Agriculture Organization (FAO), OCHA, the Development Programme (UNDP), the Population Fund (UNFPA), the Human Settlements Programme (UNHABITAT), the UNHCR, the Children's Fund (UNICEF), the World Food Programme (WFP), and the WHO. About the Inter-Agency Standing Committee, INTER-AGENCY STANDING COMM., http://www.humanitarianinfo.org/iasc/pageloader.aspx?page=content-about-default (last visited Dec. 26, 2013).

27.

See generally Inter-Agency Standing Comm. [IASC], Terms of Reference for the Humanitarian Coordinator (May 4, 2009) (endorsed by 73rd IASC Working Group) [hereinafter the role and responsibilities of the humanitarian coordinator).

IASC, HC TOR]

(describing

28.

Strengthening the Coordination of Humanitarian Emergency Assistance of 34, U.N. Doc. A/RES/46/182 (Dec. 19,

the United Nations, G.A. Res. 46/182, Annex, 1991).

29.

Jan Egeland, Towards a Stronger Humanitarian Response System, 24

FORCED MIGRATION REV., IDP SuPP. 4, 4 (2005).

forth the aims of the cluster approach).

31. 32. 33. 34. 35.

Egeland, supra note 29, at 4. HUMANITARIAN RESPONSE REVIEW, supra note 19, at 51. Egeland, supra note 29, at 4. IASC, CLUSTER APPROACH, supra note 20, at 2. Id. at 2-4.

are

"adequate

tasked with ensuring

clusters are assembled as needed when disaster strikes, and lead agencies coordination mechanisms ... adequate preparedness, as well as adequate strategic planning."36 Country clusters are often subdivided by geographic scale, establishing national-level clusters the capital and subnational clusters at the provincial or local level.37 The lead agency of a local cluster is determined by the UN official overseeing operations-the humanitarian coordinator-and it is not necessarily identical to the lead agency at the global level.38

Clusters are to be activated in all major emergencies, the existence of which is judged by the scale of the needs and by the complexity of the response.39 In principle, the humanitarian coordinator establishes the clusters and selects cluster leads at the earliest possible opportunity-on paper this is within forty-eight hours.40 Lead agencies should be chosen based upon "existing operations and capacities" and after consultations among the agencies operating at country level (see Table 1).41 A practice has developed of appointing NGOs to co-chair clusters alongside a UN agency, in order to reduce UN dominance of the cluster system, prevent conflicts of interest, and generally legitimize the system in the eyes of NGOs.42 IASC policy requires that membership within the

36. 37.

Id. at 10. INTER-AGENCY STANDING COMM. [IASC], HANDBOOK FOR RCs AND HCs ON IASC,

(2010)

EMERGENCY PREPAREDNESS AND RESPONSE 35-39 HANDBOOK]. 38.

IASC, CLUSTER APPROACH, supra note 20, at 11-13. The guidance document explaining the approach does express a preference that the global cluster lead be named a lead agency at the national level. Id. at 5 ("To enhance predictability, where possible sector lead arrangements at the country level should be in line with the lead agency arrangements at the global level."); see also INTER-AGENCY STANDING CoMM. [IASC], OPERATIONAL GUIDANCE ON DESIGNATING SECTOR/CLUSTER LEADS IN MAJOR NEW EMERGENCIES, at 2, REF 1.4.4 (May 23, 2007) [hereinafter IASC, DESIGNATING CLUSTER LEADS] (stressing flexibility in selecting lead agencies).

39.

IASC, DESIGNATING CLUSTER LEADS, supra note 38, at 1 ("For IASC operational purposes, a 'major new emergency' is defined as any situation where humanitarian needs are of a sufficiently large scale and complexity that significant external assistance and resources are required, and where a multi-sectoral response is needed with the engagement of a wide range of international humanitarian actors.").

40.

The IASC provides for a six-step standard operating procedure

for designating cluster leads: (1) consultations with local government, UN agencies, NGOs, and other international organizations to determine capacities, leaders, cross-cutting issues, and needed OCHA support; (2) a proposal is drafted by the humanitarian coordinator and forwarded to New York; (3) the head of OCHA reviews the proposal with the members of the IASC; (4) the OCHA head ensures that the IASC agrees at the global level; (5) OCHA informs the in-country coordinator of its decision; and (6) the coordinator informs local government and country-level partners. Id. at 3. This process is designed to take forty-eight hours; it is not clear how often practice meets this expectation. Id.

41. 42.

Id. at 2. See, e.g., DOMITILLE KAUFFMANN & SUSANNA KROGER, IASC CLUSTER APPROACH EvALUATION, 2ND PHASE, COUNTRY STUDY-MYANMAR 30 (2010) (noting

clusters at country level be held open to groups with "real operational capacities" in the relevant sectors."

Table 1: Global and Country-Level Cluster Lead Agencies

Cluster and Cross-Cutting Issues

Global Cluster Lead

Cluster Lead: Haiti After 2010 Quake

Cluster Lead: Burma After Cyclone Nargis

Response areas'

Agriculture Camp Coordination/

Camp Management

Early Recovery

FAO UNHCR (conflict)

FAO IOM

& IOM (other) UNDP

UNDP

Education

Save the

Children

FAQ N/A

Resident

Coordinator

Save the Children

Emergency Shelter

UNHCR & IFRC

IFRC

UN-HABITAT

Food Delivery

Health

Nutrition

Protection

WHO

UNHCR

WHO & PAHO

WHO & Merlin

Human Rights,

UNICEF (child),

Save the Children

& UNFPA

(referred to as

(gender-based

protectin of

violence)

children and

Water and Sanitation

OCHA, WFP, &

WHO & PAHO

Telecommunications

-UNICEF

Logistics

WVFP

that NGO cochairs in Myanmar "proved very beneficial ... both in terms of improved leadership and continuity of cluster activities"). In some countries, this has failed due to unexpected resistance from NGOs. In Haiti, "[nleither NGOs nor the government (with some exceptions) wanted to co-facilitate clusters. Reasons ... included Haiti's still shaky political landscape, fear of exposure to public scrutiny and critique, and the NGOs' worry of decreasing their scope for advocacy vis-h-vis the United Nations." ANDREA BINDER & FRANCOIS GRfJNEWALD, IASC CLUSTER APPROACH EVALUATION, 2ND PHASE: COUNTRY STuDYUHAITI 30 (2010).

43.

IASC, DESIGNATING CLUSTER LEADS, supra note 38, at 3.

Cluster and Cross-Cutting Issues

Global Cluster Lead

Cluster Lead: Haiti After 2010

Cluster Lead: Burma After

Environment

Gender

UNEP Cochairs of IASC

N/A (but note the

N/A (but note the

Quake

Subworking

role of gender-

role of gender-

Group on Gender,

based violence in

based violence in

in 2012:

the protection

the protection

clusters)

clusters)

UNICEF, UNHCR, Relief Int'l, & Int'l

UNAIDS

N/A (see health)

N/A (see health) HIV/AIDS Abbreviations (from top left): Food and Agriculture Organization (FAO), UN High Commissioner for Refugees (UNHCR), International Organization for Migration (IOM), UN Development Programme (UNDP), UN Children's Fund (UNICEF), International Federation of Red Cross and Red Crescent Societies (IFRC), UN Human Settlements Programme (UN-HABITAT), UN Population Fund (UNFPA), Office for the Coordination of Humanitarian Affairs (OCHA), World Food Programme (WFP), World Health Organization (WHO), Pan-American Health Organization (PAHO), UN Environment Programme (UNEP), Inter-Agency Standing Committee (IASC), Joint UN Programme on HIV/AIDS (UNAIDS).

HUMANITARIAN RESPONSE (Nov. 24, 2006); Sub- Working Group on Gender in

Humanitarian Action, INTER-AGENCY STANDING COMM., http://www.humanitarianinfo.org/iasc/pageloader.aspx?page=content-subsidi-

tf gender-default; Haiti, HuMANITARIANRESPONSE, https://haiti.humanitarianresponse.info/ (last visited Jan. 5, 2014).

Although the cluster approach does not create any new legal relationships between lead agencies and other humanitarian actors, it imposes on lead agencies the duty to ensure a range of conditions.44 In particular, cluster leads are responsible for ensuring inclusion of key partners, establishment humanitarian coordination mechanisms," and interaction with national actors, including state officials.45 Moreover, clusters "should ensure adherence to norms, policies and standards agreed at the global

"appropriate

44.

45.

IASC, CLUSTER APPROACH, supra note 20, at 7.

lead organizations also

level."46 Cluster the ultimate responsibility for ensuring that so-called crosscutting issues, such as gender, are properly taken into account,47 a problem that has plagued many responses.48 Finally, IASC documents state that the cluster lead is intended to take a representative role, acting as a voice for the interests of organizations operating within its sector in meetings with higher level UN officials and other cluster leads.49

retain

include

Cluster activities largely take place through regular meetings of participants.50 The goals of such meetings information sharing, feedback, coordination on strategy and activities, and the preparation of joint funding appeals.51 A recent evaluation notes, "As such, clusters act as platforms for achieving coordination, as well as elements that go beyond mere coordination, such as peer review, learning, or the organization of a common response ... ."52 Similar meetings took place under older coordination arrangements, but the cluster approach to these models by identifying clear leadership, connecting the in-country meetings to a global framework, and designating cluster leads as "providers of last resort."53

improve on

is intended

The latter concept-provider of last resort-remains notoriously unclear and underused, and it constitutes a barrier to the full implementation of the cluster approach.54 Where "critical gaps" appear in the response, the cluster lead is required to either convince a partner to address the problem, provide the service directly, or work with the UN representative and donors to obtain further funding and resources.55 Reviewers have noted that this concept has generally

46. 47.

Id. at 6. See id. at 7 (noting this as a "particular responsibility" of lead agencies); IASC, DESIGNATING CLUSTER LEADS, supra note 38, at 1 (highlighting this responsibility and providing a partial list of such cross-cutting issues).

48. 49.

STEETS ET AL., supra note 10, at 56-58. IASC, Cluster Lead TOR, supra note 44, at 1 (describing the lead's role in advocating to and coordinating with local and international authorities). Lead agencies also serve as members of the UN "Humanitarian Country Team," which also includes the UN humanitarian coordinator and national authorities if possible. IASC, HANDBOOK, supra note 37, at 35-36.

50.

STEETS ET AL., supra note 10, at 24 ("These clusters usually meet regularly--on a daily, weekly, monthly or quarterly basis, depending on the intensity of the crisis ....

).

51.

Id. at 24-25 ("These clusters ... share information and provide mutual feedback among members, create cluster strategies and work plans, contribute to the preparation of major funding appeals, such as the Common Appeals Process (CAP), or organize joint activities.").

52. 53. 54. 55.

Id. at 25. Id. See, e.g., id. at 75-76 (urging participants not to "give up" on the concept). IASC, CLUSTER APPROACH, supra note 20, at 10; INTER-AGENCY STANDING COMM., OPERATIONAL GUIDANCE ON THE CONCEPT OF 'PROVIDER OF LAST RESORT' (June 20, 2008), available at http://www.refworld.org/docid/4986da912.html

VANDERBILTJOURNAL OF TRANSNAT1ONAL LAW

failed to ensure that such gaps are addressed,56 but, in recent responses, the concept appears to have been employed more readily.57 To the extent the concept of provider of last resort is operationalized, the choice of cluster leads becomes all the more significant.

The framework for accountability established by the cluster approach is examined in a later Part.58 But it must be noted at the outset that ruptures in the accountability chain will not necessarily or even generally lead to a system-wide breakdown. In response to a deficit in central leadership, actors tend to feel less accountable to the humanitarian coordinator,59 but the agencies within the cluster system remain active and, importantly, retain their status as cluster leads.60 In this environment, it is possible that certain clusters will be able to maintain influence despite the breakdown in authority at higher levels. When this happens, one may think of the cluster system less as consolidating authority within the central actors of the OCHA system but rather leadership capacities on the cluster leads themselves, potentially without the accompanying accountability structures.

as bestowing additional

IASC, GUIDANCE ON LAST RESORT] (affirming the responsibility of cluster leads to fill critical gaps or, where resources fail, work with the national government, the United Nations, and donors to ensure an effective response).

56.

STEETS ET AL., supra note 10, at 56 ("[O]nly 26% of survey respondents indicated that they had experienced situations in which a cluster lead agency had acted as provider of last resort."); INT'L COUNCIL OF VOLUNTARY AGENCIES [ICVA], THE ROLLOUT OF THE CLUSTER APPROACH IN THE DEMOCRATIC REPUBLIC OF CONGO (DRC) (Mar. 18, available at https://icvanetwork.org/ doc00001846.html (noting that providers "were not stepping forward quickly enough").

ICVA,

DRC],

2006)

57.

See, e.g., INTER-AGENCY STANDING COMM.

[IASC], RESPONSE TO THE HUMANITARIAN CRISIS IN HAITI FOLLOWING THE 12 JANUARY 2010 EARTHQUAKE: ACHIEVEMENTS, CHALLENGES AND LESSONS TO BE LEARNED12 n.57 (2010) (noting that IOM acted as the "manager of last resort" in at least one internally displaced person camp); RAJ RANA & JEREMY CONDOR, INT'L ORG. FOR MIGRATION [IOM], EVALUATION OF THE INTERNATIONAL ORGANIZATION FOR MIGRATION's ONGOING ACTIVITIES ON SUPPORT TO THE FLASH APPEAL FOR THE HAITI EARTHQUAKE AND CHOLERA OUTBREAK 13 (2011) (stating that the IOM's work with water and sanitation in the camps has "reinforced IOM's credibility as the CCCM Cluster lead, [and] committed the organization to its role as the provider of last resort. This is a commitment rarely implemented by other Cluster leads, and has to be strongly applauded."). In September of 2010, the IOM was "acting as a Camp Management Agency ... in 120 camps around Port-au-Prince" and elsewhere." Camp Management Operations Unit [CMO], Int'l Org. for Migration & Camp Coordination Camp Management Cluster, CMO Weekly Report, 117-17 September: CMO IN HAITI, http://cmohaiti.wordpress.com/ (ast visited Dec. 26, 2013).

in Action, CAMP MANAGEMENT OPERATIONS

58. 59.

See infra Part V. See, e.g., KAUFFMANN & KROGER, supra note 42, at 38-39 (noting that

actors did not feel accountable to the humanitarian coordinator).

60.

Cf. id. at 39 (noting the rise of informal accountability structures within

clusters).

III. THE POWER OF HUmANITARIAN COORDINATION: OPERATIONAL AND INSTITUTIONAL ASPECTS

It should be clear that the cluster system does not attempt to wholly unify humanitarian policy, either globally or at the country level. Still, in navigating the tension between collaboration and diversity, the UN framework creates some important opportunities for the exercise of power. For example, the clusters and the IASC itself adopt standards, best practices, and operational guidelines that may affect the accepted range of practices in future emergencies.61 The system might also be examined as a force that legitimizes the "mission creep" of some organizations into areas not expressly covered by their mandates.62 This Article cannot possibly cover all aspects of the UN framework, and there is much left to explore.

This Part focuses on the potential problems associated with the system's emphasis on leadership and on appointing lead agencies. This mechanism resembles a form of institutional choice, which inevitably affects the way that a given problem will be addressed.63 First, the relative influence of leaders over policy and funding risks pathologically harmonizing policy for experimentation and competition. Second, by selecting a lead agency to manage the response in each sector, the cluster approach exercises a form of institutional choice that magnifies the power of a particular agency and brings that agency's unique practices and principles to bear on an affected population. Each aspect pulls in a different direction, thus emphasizing the problem of ensuring an appropriate

allowing

rather

than

61.

As with most forms of international standard setting, the proliferation of best practices may serve to legitimize certain methods while freezing out organizations that take alternative approaches. See, e.g., Benedict Kingsbury, Operational Policies of International Institutions As Part of the Law-Making Process: The World Bank and Indigenous Peoples, in THE REALITY OF INTERNATIONAL LAW: ESSAYS IN HONOUR OF IAN BROWNLIE 323, 324-25 (Guy S. Goodwin-Gil & Stefan Talmon eds., 1999) (discussing efforts to include the voices of indigenous peoples in policymaking at the World Bank).

62.

See ERNST B. HAAS, WHEN KNOWLEDGE IS POWER: THREE MODELS OF the ability of

CHANGE international organizations to draw jurisdictional boundaries in complex operations).

INTERNATIONAL ORGANIZATIONS

(noting

(1990)

63.

The concept of "institutional choice" is developed most fully in other contexts. See generally NEIL K. KOMESAR, IMPERFECT ALTERNATIVES: CHOOSING INSTITUTIONS IN LAW, ECONOMICS, AND PUBLIC POLICY 3-13 (1994) (developing a "participation-centered approach" to comparative institutional analysis). Gregory Shaffer and Joel Trachtman have recently discussed institutional choice in their analysis of WTO law, noting that the design of legal structures and the interpretation of rules may direct decisions to different "social decision-making processes," or institutions. Gregory Shaffer & Joel Trachtman, Interpretation and Institutional Choice at the WTO, 52 VA. J. INT'L L. 103, 105 (2011). This Article is sympathetic to this definition insofar as it requires a consideration of the relative merits of the various institutions to which decision-making power might be directed, but in turning to the accountability and legitimation of this power in subsequent sections, this Article veers away from the type of analysis that Shaffer and Trachtman undertake.

254.

VANDERBILT]OURNAL OF TRANSNATIONAL LAW

balance between shared policy and legitimate difference.64 These problems will of accountability and legitimacy.

investigating questions

the stage

set

for

A. Trampling Diversity? Harmonization of Policy in the Clusters

Because the cluster approach is intended to coordinate policies among various actors working in the same sector, its very nature gives rise to fears that alternative views will be abandoned, marginalized, or co-opted.65 In addition, the system might magnify the influence of certain parties. The inclusion of development actors, peacekeeping forces, military personnel, or governments in cluster meetings may be important from the perspective of increased coordination. But, to the extent these actors influence policies, this practice threatens humanitarian principles of neutrality and independence.66 This subpart addresses these concerns through a handful of examples drawn from the life of the cluster system.

To understand the delicate balancing act between harmonization the wide diversity among and difference, one must appreciate humanitarian actors. Over the decades, aid agencies have developed divergent policies toward peacemaking, human rights, intervention, civil-military economic development, and anything else that might fall under the heading of politics. For now it suffices to note that many believe diversity in relief policy to be advantageous, and thus, the pressure to coalesce around a single strategic plan raises some concerns.67

long-term

relations,

50 (2004).

See, e.g., HUMANITARIAN POLICY GRP., HPG BRIEFING NOTE: THE CURRENCY OF HUMANITARIAN REFORM 6 (2005) ("The cluster lead approach also risks reinforcing existing stereotyped responses, and failing to take account of the interests of beneficiaries in the particular context.").

See, e.g., STEERING COMM. FOR HUMANITARIAN RESPONSE [SCHR], SCHR POSITION PAPER ON HUMANITARIAN-MILITARY RELATIONS 11 (2010) [hereinafter SCHR, POSITION PAPER] ("Military presence in cluster meetings (in particular protection) may inhibit the free exchange of information amongst humanitarian organisations, and give rise to the perception of a common strategy and objectives between humanitarians and the military."); STEETS ET AL., supra note 10, at 64 ("When it is no longer voluntary, closer cooperation threatens to undermine independence, neutrality and impartiality for example when humanitarian organizations are financially dependent on clusters or their lead agencies, or when clusters or their leads have close links to integrated missions, peacekeeping operations, governments or other actors that are parties to the conflict."). 67.

The Global Humanitarian Platform, a group of UN and non-UN humanitarian organizations, issued a statement on partnership and coherence in 2007 that emphasizes principles of equality among organizations, transparency, resultsbased action, responsibility, and complementarity. On the latter, the statement notes, "The diversity of the humanitarian community is an asset if we build on our comparative advantages and complement each other's contributions." Press Release,

25S

Standardization in the cluster system may work through a number of different dynamics. Understanding the cluster approach as a network of like-minded actors working for a common purpose raises the possibility that participants will be socialized through repeated into taking a common position.68 'In his study of interaction transgovernmental networks, Kal Raustiala identifies an alternative rational-choice account, whereby network arrangements increase the benefits of and incentives for policy convergence.69 Both of these dynamics may be reflected in the cluster system-one recent survey refers to the "authority of format" imposed by the cluster approach.70 But the most important harmonizing effects relate to the funding of humanitarian activities.

In some cases, donors have enforced conformity

to policies developed within certain clusters. For example, a Mddecins sans Frontidres (MSF) review of cluster operations noted that, in Uganda, the European Community Humanitarian Office "required 'partners' to fit their proposals into existing cluster strategies before granting funding."7' And there is evidence that donors continue to. view participation in funding decisions,72 indicating that active engagement with the clusters may bring tangible benefits. In addition, the cluster system may be changing donors' funding habits, as donors begin to delegate allocation decisions to humanitarian coordinators and cluster

in clusters as an

important aspect

Global Humanitarian Platform, Principles of Partnership (July 12, 2007), available at http://www.globalhumanitarianplatform.org/pop.html#pop.

68.

See SLAUGHTER, supra note 64, at 198-200 (explaining how socialization operates on international government networks). See generally Jeffrey T. Checkel, International Institutions and Socialization in Europe: Introduction and Framework, 59 INT'L ORG. 801 (2005) (exploring the socializing role of institutions in Europe and the implications of that role).

69.

Kal Raustiala,

The Architecture of International Cooperation: Transgovernmental Networks and the Future of International Law, 43 VA. J. INT'L L. 1, 62-68 (2002). This understanding draws on Mark A. Lemley & David McGowan, Legal Implications of Network Economic Effects, 86 CA. L. REV. 479 (1998).

70. Matthew Serventy, National NGOs and the Cluster Approach: The Authority of Format, HUMANITARIAN EXCH., Jan. 2013, at 33-35 ("The cluster approach in itself is a format, and some activities may not be undertaken simply because they do not fit neatly into the cluster structure; as one Somali NGO put it: 'we have learnt to structure our programmes the way clusters are structured'.").

71.

Katharine Derderian et al., UN Humanitarian Reforms: A View from the Field, HUMANITARIAN EXCH., June 2008, at 36, 37, available at http://www.odihpn.org/ humanitarian-exchange-magazine/issue-39/un-humanitarian-reforms-a-view-from-thefield.

72.

See, e.g., U.K. DEP'T FOR INT'L DEV. [DFID], MULTILATERAL AID REVIEW: ENSURING MAXIMUM VALUE FOR MONEY FOR UK AID THROUGH MULTILATERAL ORGANISATIONS http://www.dfid.gov.uk/Documents/ publications1/mar/multilateralaidreview.pdf (stating that organizations are to be evaluated in part on their "partnership behaviour" and coordination functions).

available at

(2011),

leaders.73 The fact that OCHA has warned against this practice only demonstrates that donors and other external actors may take advantage of the cluster approach in ways its designers did not anticipate.74

The clusters have also

integrated with other institutional mechanisms at the United Nations in ways that encourage the harmonization of policy. OCHA manages a "Consolidated Appeals Process" for humanitarian financing, through which cluster lead agencies, in consultation with other NGOs and agencies, develop strategic action plans in an integrated, rather than competitive, manner.75 By its very nature, this process dissuades competition and encourages the creation of common plans and operational standards, though it can also allow more powerful or influential organizations to trample others in the priority-setting process.76 Giving clusters substantial authority over the distribution of common, or pooled, funds, empowers them to implement common action plans, but this tends to "silence peer criticism" and leads to "horse-trading" among participating agencies that tends to decrease the quality of aid.77 Finally, the cluster approach gives NGOs the opportunity to partner with UN agencies and the IOM, which have exclusive access to certain funds.78 This is a positive trend for many organizations, but it necessarily privileges programs that are undertaken in concert with the United Nations, as opposed to those undertaken independently.79

73.

Oxfam Int'l, Missing Pieces? Assessing the Impact of Humanitarian Reform available at

(Oxfam Briefing Paper, Oct. in Pakistan 7 www.oxfamamerica.org/files/oxfam-pakistan-paper-missing-pieces-01- 10-09.pdf (arguing that this arrangement is inappropriate and premature, adding "additional layers of time and bureaucracy to the disbursement" of funds).

2009),

1,

74. 75.

See id. (noting OCHA's objections). See About the CAP: About the Process, UNITED NATIONS OFFICE FOR THE COORDINATION OF HUMANITARIAN AFFAIRS, http://www.unocha.org/cap/about-thecap/about-process (last visited Dec. 26, 2013).

76.

smaller and more

See STEETS ET AL., supra note 10, at 40 (noting that "donor preferences in CAP and Flash Appeals that typically favor large result in funding trends to international organizations over disappointment among many NGOs and reinforces their financial dependence on UN agencies."). 77.

Id.; see BINDER & GRONEWALD, supra note 42, at 33-39, 52 ("This increased connectedness and ownership of the cluster, but sometimes also slowed activities, since much lobbying was necessary."); see also Chad: Mixed Verdicts on Coordination of Massive Relief Effort, IRIN AFRICA (Jan. 23, 2008), http://www.irinnews.org/report/ (reporting the opinion 76386/chad-mixed-verdicts-on-coordination-of-massive-relief-effort of an International Rescue Committee representative, who stated that clusters are "first and foremost about politics and money").

local ones. This leads

78.

STEETS ET AL., supra note 10, at 40 (pointing to CERF, which is off-limits to NGOs). On this fund, see CENTRAL EMERGENCY RESPONSE FUNDS, CERF FACTS 2 (Dec. 2011), available at http://ochanet.unocha.org/plDocuments/FactSheetAH.pdf (describing who can apply for CERF funds).

79.

See Serventy, supra note 70, at 34 ("As always money equals power, and

the funding power stays firmly in the hands of international agencies.").

In addition to matters relating to funding, activities within clusters may become standardized through the adoption of certain performance measures. In Chad, the Water Sanitation Hygiene (WASH) cluster employed Sphere Project standards for the provision of services.80 The Sphere Project provides a set of indicators, which purport to be based on international human rights standards, providing for minimum quantities of clean drinking water and other services.8' Because the indicators as set at the global level tended to strain natural resources and exacerbate local conflicts, some NGOs pressed for these indicators to be adapted and changed to fit the Chadian context, a modification resisted by the WASH cluster lead.82 This experience may be especially problematic not only because Sphere is used to hold NGOs accountable to their donors and thus may work substantial influence over the response,83 but also because it shows the failure of the cluster system to develop responsive, context-sensitive metrics for performance evaluation.

In the early days of the cluster approach, the system was also criticized for subordinating aid to other political goals. The MSF review notes several examples where cluster funds were used primarily to further the political goals of the peace-building efforts, including in C6te d'Ivoire, where the overwhelming majority of Central Emergency Response Fund (CERF) funds were used in and around a town that suffered from anti-UN riots.84 Other aid groups have noted the risk that information shared in clusters will be appropriated and used by the military, thus risking the neutrality of humanitarian actors and potentially frustrating access.85

Harmonization is not a danger in all cases, and sometimes the fear that policies will become overly rigid or standardized is more theoretical than real. In the response to the 2008 hurricanes and tropical storms in Haiti, for example, the response experienced the opposite problem, as few common plans or strategies were implemented.86 But where the system exerts substantial influence

80.

FRANCoIS GRONEWALD & BONAVENTURE

SOKPOH,

JASC CLUSTER

APPROACH EVALUATION, SECOND PHASE COUNTRY STUDY: CHAD 31 (2010).

81.

THE SPHERE PROJECT, WHAT IS SPHERE?, HUMANITARIAN CHARTER AND MINIMUM STANDARDS IN HUMANITARIAN RESPONSE (3d ed. 2011) [hereinafter SPHERE HANDBOOK]. For problems in the development and application of Sphere indicators, see Margaret L. Satterthwaite, Indicators in Crisis: Rights-Based Humanitarian Indicators in Post-Earthquake Haiti, 43 N.Y.U. J. INT'L L. & POL. 865 (2011).

82. 83.

GRONEWALD & SOKPOH, supra note 80, at 31. See Satterthwaite, supra note 81, at 963 ("[O]nce codified ...the debates underlying what it means to provide quality assistance ... tend to retreat from view.").

84. 85.

Derderian et al., supra note 71, at 37. See generally SCHR, POSITION PAPER, supra note 66, at 4-7, 11 (discussing

policies governing the relationship between humanitarians and armed forces).

86.

BINDER & GRONEWALD, supra note 42, at 30.

either directly or via pressure from donors, its ability to trample legitimate diversity will become a cause for concern.

B. Problematizing Diversity: Variance Among Leaders

real effect on

The diversity of participating actors in the cluster approach can also become a problem through the leadership functions of the system. In setting global leaders and country-level lead agencies, the cluster system engages in a process of institutional choice that works a institutional structures, budgets, and, most importantly, affected populations. Humanitarian organizations take a range of divergent approaches to neutrality and independence, or to idea of principled action.87 Organizational cultures, the entire funding structures, and competences add additional dimensions to the diversity among actors.88

As a facilitator and coordinator of cluster meetings, a lead agency may wield considerable power to exclude or include certain actors. This dynamic is currently playing out in Haiti, where a group of human rights advocates, acting on behalf of women and girls residing in twenty-two different camps for displaced persons, have argued that the gender-based violence subcluster "refuses to include Haitian grassroots women's groups to meaningful[1y] participate in the planning and implementation of activities designed to address sexual violence . ."89 This same coalition has succeeded in obtaining precautionary measures against Haiti from the Inter-American Commission for Human Rights, which called on the country in 2010 to ensure, inter alia, that grassroots organizations participate effectively in coordination arrangements.90

87.

See generally Nicholas Leader, The Politics of Principle: The Principles of Humanitarian Action in Practice (Humanitarian Pol'y Grp., Report No. 2, Mar. 2000) (analyzing shifts in contemporary approaches to humanitarian principles); Kate Mackintosh, The Principles of Humanitarian Action in International Humanitarian Law (Humanitarian Pol'y Grp., Report No. 5, Mar. 2000).

88.

(Apr.

For a snapshot of the relative skills of different agencies working in Haiti, see Vince Beiser, Organizing Armageddon: What We Learned from the Haiti Earthquake, WIRED http://www.wired.com/magazine/2010/04/ ff_haitilall/1. 89.

MADRE et al., Gender-Based Violence Against Haitian Women & Girls in Internal Displacement Camps: Submission to the United Nations Universal Periodic Review, Republic of Haiti, 1] 20, 12th Sess., Working Group on the Universal Periodic Review, Hum. Rts. Council, Oct. 3-14, 2011, available at http://ijdh.org/wordpress/wpcontent/uploads/2011/03/UPR-GBV-Final-4-4-2011.pdf.

2010),

19,

90.

Audio tape: Women and Girls in Camps for Forcibly Displaced Persons in Haiti, PM 340/10, held by Inter-American Comm'n on Human Rights (Mar. 22, 2011), available at http://www.oas.orgles/cidh/audiencias/topicslist.aspx?lang-en&topic=23. Note that, in this dispute, the victims' advocates advance the novel argument that the subcluster constitutes an "agent" of the state under the Articles on State Responsibility, Articles on the Responsibility of States for Internationally Wrongful Acts, G.A. Res. 56/83, Annex, Art. 5, U.N. Doc. AIRES/56/83 (Jan. 28, 2002), and that the country is thus responsible for its failures. The Commission granted precautionary

is designated

to fill any operational gaps,

The choice of a cluster lead takes on an operational dimension through the lead agency's role as a provider of last resort.91 Where a single agency its institutional practices and principles will be that much more likely to inform conditions on the ground. This can be particularly important where and needs-based treatment is concerned,92 as some organizations, particularly the UNHCR, are committed by to a status-based approach.93 While few disasters have seen cluster leads asserting this role,94 it is possible that the 2010 Haiti earthquake and other examples will herald an increasing assertion of the provider-of-lastresort concept.95

the difference between status-based

their mandate

The consequences of institutional choice may be expressed through the example of Camp Coordination and Camp Management (CCCM).96 Over time, camps of displaced persons can resemble sophisticated polities, requiring a range of municipal-like services governing infrastructure, governance arrangements, markets, and police.97 Agencies involved in CCCM are responsible for, among other

measures but did not appear to directly address this point, which, if affirmed, could significantly alter the dynamics of the response.

91.

For an elaboration of this concept, see supra text accompanying notes 53-

57.

92.

On this, see GRONEWALD & SOKPOH, supra note 80, at 31 ("Resource allocation methods and technical choices are different depending on which approach is chosen."); STEETS ET AL., supra note 10, at 51 ("[T]here is a conflict between the statusbased approach to humanitarian assistance for Internally Displaces Persons (IDPs) and other affected groups espoused by UNHCR and the needs- or vulnerabilities-based approach of most other humanitarian actors.").

93.

See Statute of the Office of United Nations High Commissioner for (V), Annex, 1 2 (Dec. 14, 1950) (requiring the high

Refugees, G.A. Res. 428 commissioner's work to "relate, as a rule, to groups and categories of refugees").

94.

See, e.g., KAUFFMANN & KROGER, supra note 42, at 14 (noting that lead agencies "acted as 'advisor of last resort', not as 'provider of last resort' as no financial resources were available"); SUSANNA KROGER & JULIA STEETS, IASC CLUSTER APPROACH EVALUATION, 2ND PHASE: COUNTRY STUDY-THE OCCUPIED PALESTINIAN TERRITORIES 22 (2010) (noting the same tendency).

95.

See IASC, Response to the Humanitarian Crisis in Haiti, supra note 57, at 12 n.57 (noting that the IOM acted as the "manager of last resort" in at least one internally displaced person camp). In addition, an agency's status as cluster lead in a particular field may generate greater donations for that purpose, thus increasing the agency's on-the-ground involvement without having to invoke the concept.

96.

On the general concept of CCCM, see NORWEGIAN REFUGEE COUNCIL & CAMP MGMT. PROJECT, THE CAMP MANAGEMENT TOOLKIT (2008) [hereinafter CM TOOLKIT] (providing guidelines for the management of refugee camps).

97.

Cf. Ralph Wilde, Note, Quis Custodiet Ipsos Custodes?: Why and How UNHCR Governance of "Development" Refugee Camps Should Be Subject to International Human Rights Law, 1 YALE HUM. RTS. & DEV. L.J. 107, 108 (1998) (describing this phenomenon in the case of longer term "development camps"); Mark Pallis, The Operation of UNHCR Accountability Mechanisms, 37 N.Y.U. J. INT'L L. & POL. 869, 883-86 (2005) (elaborating on Wilde's account); Satterthwaite, supra note 81, at 874 (noting that humanitarians may "govern the putatively ungoverned-those

VANDERBIL TIOURNAL OF TRANSNA TIONAL LAW

[VOL, 47-239

functions, coordinating and monitoring service delivery, maintaining infrastructure, "establishing governance and community participation/mobilisation mechanisms."98 Camp coordinators are responsible for macro-level strategy, for monitoring the management of camps, and, crucially, for developing camp closure and exit strategies.99 This latter power tension between maintaining lives and livelihoods in the camp, and the importance of identifying more permanent and sustainable solutions to mass displacement.100 Thus, CCCM entails not only the authorities of a municipal government but also the life-and-death decisions relating to the continued existence of a camp.

involves the important

The IOM, a treaty-based international organization that is formally independent from the United Nations, is a frequent cluster leader in this field.10' A relative newcomer to camp coordination and management, the IOM has at times been the target of other relief agencies and human rights groups,102 and its lack of a clear mandate for protection or humanitarian action has been cause for some concern.103 This critique should not be overstated, as the agency has formally adopted certain principled guidelines,104 and the IOM has been praised for many of its humanitarian activities.105 But it is widely acknowledged that the IOM operates with a much more pragmatic or technical orientation, placing less emphasis on broadly

whose governments have failed, become predatory, or can no longer be counted on to provide protection").

CM TOOLKIT, supra note 96, at 28. See id. at 30-31 (discussing the primary objectives of camp coordination). See id. at 208-17 (discussing aspects of camp closure and durable solutions

98. 99. 100. for refugees). 101.

The organization is a global cluster lead for CCCM in nonconflict situations. See IASC, CLUSTER APPROACH, supra note 20, at 3. It has served as a CCCM lead in at least four disasters and has led other clusters as well. INT'L OFFICE OF MIGRATION, IOM AND THE CLUSTER APPROACH (IOM Doc. 05/EPC-CLUSTER/O807), available at http://www.iom.intijahia/webdav/shared/shared/mainsite/activities/mepmm/op support/epc-clusters_070808.pdf.

102.

E.g., Tim Morris, IOM: Trespassing on Others' Humanitarian Space?, 22 FORCED MIGRATION REV. 43, 43 (2005) (discussing other organizations' criticisms of the IOM); Editorial, IOM, Darfur, and the Meaning of Undermining (MOU), 6 TALK BACK, (discussing Oct. 4, 2004, available at http://www.icva.ch/docOO001253.html#iom concerns about the IOM's expertise and capacity to accomplish its objectives).

103.

See, e.g., Melanie Teff & Emilie Parry, Haiti: Still Trapped in Emergency

Phase, REFUGEES policy/field-report/haiti-still-trapped-emergency-phase IOM).

6, 2010),

INT'L

(Oct.

http://www.refugeesinternational.org/ the

(discussing criticism of

104.

The IOM was an early adopter of the Guiding Principles on Internal Displacement, which purport to set out the relevant international legal obligations with respect to internally displaced persons. INT'L ORG. FOR MIGRATION, INTERNALLY DISPLACED PERSONS: IOM POLICY AND ACTIVITIES T 10 (IOM Doc. MC/INF/258, Nov. 18, 2002).

105.

E.g., RANA & CONDOR, supra note 57, at 22-26 (characterizing many of the IOM's efforts as successful and stating that the "IOM has conspicuously and consistently delivered on its promises").

phrased principles or theoretical discussion.106 Owing in part to its technical orientation, and in part to its status as a treaty-based international organization, the IOM has developed the view "that the organisation cannot tell governments what they should do or how to do it."'07

The IOM has weathered criticism in Haiti, where it acts as a cluster lead, for failing to stop government-led closures of displacedperson camps set up after the 2010 earthquake.108 The organization's alleged cooperation in some eviction proceedings raises the criticism that the organization is legitimizing these evictions.109 In 2011, newly elected Haitian President Michel Martelly enlisted the IOM in his program, the "16/6 Plan," to close six camps within his first one hundred days in office.110 The program has since been criticized as an unsustainable effort that has not provided displaced residents with sufficient resources to find livable housing elsewhere."' As of this

106.

See, e.g., Elizabeth Farris & Sara Ferro-Ribeiro, Protecting People in Cities: The Disturbing Case of Haiti, DISASTERS (forthcoming) (emphasizing the differences between the UNHCR and the IOM in approaches to the protection of displaced persons). 107.

ANDERS OLIN, LARS FLORIN & BJORN BENGTSSON, SWEDISH INST. FOR PUB. ITS

ADMIN., STUDY OF THE INTERNATIONAL ORGANIZATION FOR MIGRATION AND HUMANITARIAN ASSISTANCE 22 (2008) (emphasis removed).

108.

See Ellie Happel, 'Returning to Zero: Forced Evictions in Haiti's Displaced Persons Camps, CAN. HAITI ACTION NETWORK (Mar. 1, 2012), http://canadahaitiaction.ca/ (describing the content/returning-zero-forced-evictions-haitis-displaced-persons-camps problem of forced evictions of displaced Haitians).

109.

See id. (arguing that the IOM "merely assists the relocation of camp residents to equally poor conditions" and that the IOM "reinforces the power disparity between landowners and the displaced" by failing to include residents in negotiations with the government regarding the closure of camps); see Mark Snyder, IOM's Direct Participation in Forced Evictions Raises Many Questions, CAN. HAITI ACTION NETWORK (Apr. forced-evictions-raises-many-questions families).

2012), http://www.canadahaitiaction.calcontent/ioms-direct-participation-

(questioning the impact of the expulsion of

5,

A number of sources focus on the relocation of more than a thousand displaced persons from a camp thought to be dangerous into a "barren, windswept" valley known as Corail Cesselesse. Sontag, supra note 2, at Al (noting that "some disaster experts consider [the move] to have been a mistake, imposed on a group without options"); Displaced Haitians: "We Can't Continue in This Situation Anymore", DEMOCRACY NOW! (July 12, 2010), http://www.democracynow.org/2010/7/12/displaced haitianswecant_ continuein (noting that the camp population was originally told that this would be a temporary situation). In another notorious case, after attempting to relocate the residents of one camp, IOM employees allegedly deposited a number of displaced persons at the Delmas police station. Justin Podur, The Eviction of Barbancourt 17, ZNET (Oct. 5, 2011), http://www.zcommunications.org/the-eviction-of-barbancourt-17by-justin-podur.

110.

Nelson A. King, Launch of Ambitious Housing Plan for Haiti, CARIBBEAN http://www.caribbeanlifenews.com/stories/2011/10/201 1-

2011),

(Oct.

LIFE 10_10 nk martelly.html.

12,

111.

See Kevin Edmonds, Unsustainable Solutions to Haiti's Housing Crisis, N. (July 20, 2012), http://nacla.org/blog/ AMERICAN CONG. ON LATIN AMERICA 2012/7/20/unsustainable-solutions-haiti%E2%80%99s-housing-crisis (arguing that the

Article's publication, a coalition of human rights advocates has sought preliminary measures from the Inter-American Court of Human requesting a moratorium on evictions until a more Rights, sustainable plan is developed.112

It is not clear what other agencies would have done differently if placed in the position of the IOM. Many of the problems experienced by internally displaced person camps in Haiti are likely the result of the complexity of the response, the enormity of the disaster, and the longstanding problems with corruption and land-rights issues in the country, which certainly predate the earthquake. But the very fact that another institution might have acted differently suggests two things. First, it suggests that institutional choice can work real effects on the lives of disaster victims. Second, it raises a further question: if this system can install leaders, to what extent must it then monitor their performance, correct for their mistakes, or "fire" them if things go badly?

IV. ACCOUNTABILITY, LEGITIMACY, AND INSTITUTIONAL DESIGN

In confronting its own power, the humanitarian profession has increasingly engaged in a wide-ranging and often confused discourse of accountability.113 Without firm conceptual underpinnings, any attempt to address the power of humanitarian agencies risks being swept away in the conflicting terminology of practitioners and observers, which often conceals as much as it clarifies. Without losing sight of the real-world problems identified in the prior Parts, this Part provides these concepts with some theoretical backing. With a stronger understanding of accountability as a persistent institutional problem,114 the shifting approaches to accountability carry broader

16/6 Plan has succeeded in clearing some major camps, but only at the cost of creating a more diffuse housing problem); Haiti's Housing Crisis: Human Rights Investigation Finds Forty-One Percent of Families Relocated under Haitian Government's Housing Program Live in Worse Conditions than Before Earthquake, UNDER TENTS (July 27, 2012), http://undertentshaiti.com/?p=227.

112.

Letter from Inst. for Justice & Democracy in Haiti, to Inter-Am. Comm'n on

Hum. Rts. (Feb. 27, 2012), available at http://ijdh.org/archives/25449.

ACCOUNTABILITY:

See generally NGO

POLITICS, PRINCIPLES

AND 113. (Lisa Jordan & Peter van Tuiji eds., 2006); Michael Barnett, INNOVATIONS Humanitarianism Transformed, 3 PERSP. ON POL. 723, 730 (2005) (discussing the transformation of accountability in the humanitarian field); U.N. Non-Governmental Liaison Service, NGLS Development Dossier: Debating NGO Accountability, U.N. Doc. (Aug. 2006) (by Jem Bendell) (providing a comprehensive UNCTAD/NGLS/2006/1 review of the current state of accountability while pointing to deficits and discussing recommendations for improvement); Humanitarian Accountability, 52 HUMANITARIAN EXCH. MAG., Oct. 2011, available at www.humanitarianexchange052.pdf.

114.

For an effective approach that understands accountability as a "question

rather than a clear and unequivocal goal," see Janice Gross Stein, Humanitarian Organizations: Accountable-Why,

for What, and How?,

to Whom,

lessons regarding the legitimation of humanitarian enterprises and solve old normative challenges even as they create new ones.

A. A Broad Approach to the Accountability of Networks

In an effort

to best mirror

the wide-ranging uses

in humanitarian practice, this subpart takes a broad approach to the concept of accountability. Following Ruth W. Grant and Robert 0. Keohane, this Article understands accountability to mean "that some actors have the right to hold other actors to a set of standards, to judge whether they have fulfilled their responsibilities in light of these standards, and to impose sanctions if they determine that these responsibilities have not been met.""i5 This definition, as employed by to what constitutes a its authors, takes an inclusive approach to judge," sanction, as well as to what constitutes a "right ... including a range of market and reputational mechanisms within this definition.116 Nonetheless, Grant and Keohane's view might yet be too restrictive unless the authors' reference to "a set of standards" is understood broadly, to allow for situations where no clear standards are held intersubjectively, or when those standards are in flux."x7

threatening

This broad definition has been criticized as pitching too big a that make tent, control accountability mechanisms."i8 But the broad definition is helpful for the present

the essential features

among various

institutional

unique

lose

HUMANITARIANISM IN QUESTION: POLITICS, POWER, ETHICS 124, 125 (Michael Barnett & Thomas G. Weiss eds., 2008).

115. Ruth W. Grant & Robert 0. Keohane, Accountability and Abuses of Power

in World Politics, 99 AM. POL. SCI. REV. 29, 29 (2005).

116.

Id. Other writers have adopted an inclusive approach

to defining accountability, with various degrees of enthusiasm. See, e.g., CAROL HARLOW & RICHARD RAWLINGS, LAW AND ADMINISTRATION 306-07 (3d ed. 2009) (recognizing that network forms may permit a form of accountability, though noting this is harder to secure); Jerry L. Mashaw, Accountability and Institutional Design: Some Thoughts on the Grammar of Governance, in PUBLIC ACCOUNTABILITY: DESIGNS, DILEMMAS AND EXPERIENCES 115 (Michael W. Dowdle ed., 2006) (defining accountability regimes to include not only electoral, bureaucratic, and legal mechanisms, but also market competition and means of social accountability); Aaron Bloom, Note, The Power of the Borrower: IMF Responsiveness to Emerging Market Economies, 43 N.Y.U. J. INT'L L. & POL. 767 (2011) (assessing the IMF in terms of market-type accountability).

117.

See Steve Charnovitz, Accountability of Nongovernmental Organizations (NGOs) in Global Governance 7-8 (Geo. Wash. U. L. Sch. Pub. L. & Legal Theory, 2005), available at http://scholarship.law.gwu.edul Working Paper No. 145, (criticizing Grant and cgilviewcontent.cgi?article=1428&context=faculty-publications Keohane to the extent that their definition relies on "jointly-agreed standards").

118.

These essential features are common to, inter alia, rights to sue, employeesupervisor relationships, and impeachment but do not extend to peer interactions or market dynamics. See, e.g., Richard Mulgan, Accountability': An Ever-Expanding Concept?, 78 PUB. ADMIN. 555, 566-68 (2000) (supporting a limited conception of accountability).

VANDERBIL T]OURA4L OF TRANSNA TIONAL LAW

purposes, as it enables an assessment of the competing visions of accountability facing the cluster approach. Recent reform proposals have emphasized horizontal accountability and peer review as forms of accountability.119 It is less important to second-guess the labeling of these approaches their viability as alternative institutional models. This is particularly true where, as suggested below, a discourse over accountability mechanisms can be interpreted as for legitimizing a particular system of governance.

a competition among alternative models

is to determine

than it

What is meant by the "accountability" of the cluster system is itself difficult to grasp. As noted above, the raison d'etre of the cluster approach is improving leadership and accountability across all sectors of the response.120 The primary sites for accountability seem to be the constituent NGOs and UN agencies in the system, not the system itself. But, as demonstrated above, itself becomes an institution endowed with some power, and one should rightly ask how the system may be forced to incorporate disregarded voices or to correct for any errors or misjudgments it committed when constructing a particular response. In some respects, the framing question is similar to the problem of accountability for government oversight bodies-Who watches the watchers?'2'

the cluster approach

This question is, of course, complicated by the fact that the watcher in this case is neither an individual nor a well-defined organization but is rather a loosely bound network of institutions, governments, and victims' groups interacting at and across various levels. The cluster system is thus a frame for action, as well as a potential actor in its own right.122 Though all institutions, from private firms to government departments, may be considered as

119. 120.

See infra Part VI. IASC, CLUSTER APPROACH, supra note 20, at 4 ("In the past, however, it was usually the case that only a limited number of sectors had clearly designated lead agencies accountable to the Humanitarian Coordinator. The cluster approach aims to rectify this by ensuring that within the international humanitarian response, there is a clear system of leadership and accountability for all the key sectors or areas of humanitarian activity.").

18,

2003),

(Sept.

http://www.economist.com/node/2077493

See, e.g., Who Guards the Guardians?: An Audit of Conservationists, THE 121. EcONOMIST ("Nongovernmental organisations, as many charities are pompously described these days, often escape the sort of scrutiny that they, themselves, like to apply to governments and companies."); cf. Jonathan B. Wiener & Alberto Aleemanno, Comparing Regulatory Bodies Across the Atlantic: The Office of Information and Regulatory Affairs in the U.S. and the Impact Assessment Board in the E. U., in COMPARATIVE ADMINISTRATIVE LAW 309, 312 (Susan Rose-Ackerman & Peter Lindseth eds., 2010) ("Just as regulators need oversight, so too [oversight bodies] warrant oversight ....").

122.

See generally Miles Kahler, Networked Politics: Agency, Power, and Governance, in NETWORKED POLITICS: AGENCY, POWER, AND GOVERNANCE 1-22 (Miles Kahler ed., 2009) (addressing networks through both views); KECK & SIKKINK, supra note 7, 8-10 (conceptualizing the network as a type of actor).

networks of branches, offices, and, ultimately, individuals,123 the looseness of the bonds between cluster members creates a creature that is substantially different from more traditional institutions.124 The problem is to address the accountability of such a system in a way that incorporates, but does not devolve into, the more general international the accountability of NGOs and discourse on organizations.

the system

The dual understanding of networks as actors and frameworks for action provides a useful approach. On this view, the manner in is at least partly itself is held accountable which dependent upon a latticework of relations within the organization itself.125 This view is not unknown in the contemporary judicial treatment of the firm.126 Following this approach, the trick is to ensure that this internal set of relationships can be made to work for whichever actors external to the network are deemed to be important account holders. Changing or improving accountability may consist of providing new access points to external actors, but, equally, it may to require rearranging or scrambling the internal relationships increase or decrease sensitivity to certain external or internal power centers.

B. Accountability as the Self-Justification of Institutions

The design of accountability mechanisms is interesting on its own account, as a study in the manner in which institutions are influenced and controlled. But the close tie between accountability this subject and particularly pressing for those interested in the legitimation of international law and global institutions.127 Grant and Keohane

in contemporary politics

legitimacy

renders

123.

See EDWARD L. RUBIN, BEYOND CAMELOT: RETHINKING POLITICS AND LAW FOR THE MODERN STATE 50 (2005) ("The network that constitutes the government ... is readily conceived as simply one region of a larger, more comprehensive social network.").

124.-

See generally MARY KALDOR, GLOBAL CIVIL SOCIETY: AN ANSWER TO WAR

95-97 (2003) (distinguishing networks from other organizations).

125.

For a theoretical approach that accommodates this view, see generally Francesca Bignami, From Expert Administration to Accountability Network: A New Paradigm for Comparative Administrative Law, 59 AM. J. COMP. L. 859 (2011). See also HARLOW & RAWLINGS, supra note 116, at 305-08 (examining problems that arise from multiple competing accountability relationships). For a similarly oriented critique of U.S. administrative law in the courts, see generally Elizabeth Magill & Adrian Vermeule, Allocating Power Within Agencies, 120 YALE L.J. 1032 (2011).

126.

See, for example, the cases associated with In re Caremark Int'l Inc. Derivative Litig., 698 A.2d 959, 970 (Del. Ch. 1996) (finding "a duty to attempt in good faith to assure that a corporate information and reporting system, which the board concludes is adequate').

127.

Julia Black, Constructing and Contesting Legitimacy and Accountability in Polycentric Regulatory Regimes, 2 REG. & GOVERNANCE 137, 149-50

See

describe the legitimacy as setting the background against which the accountability relationship operates: conceptions of legitimacy define the nature of the power wielder, identify the appropriate account holder, and provide the substantive norms that fuel the operation of accountability mechanisms.128 This view suggests that changes in the accountability discourse reflect more than technical adjustments, indicating deeper shifts in the underlying approaches to legitimacy in international institutions.129 This subpart proposes an approach that brings the implications of such changes to the surface.

itself to certain

This Article begins with the insight that the very rules and practices of institutions constitute arguments for an institution's legitimacy.130 By opening forms of criticism, submitting its decisions to external review, or even closing itself off from influence in the name of autonomy or independence, an institution implies a theory of its own legitimacy and opens this theory to critique. These arguments may be imposed on an institution by its masters or self-generated in an effort to secure a more stable basis for action.131 By recognizing that institutional arrangements suggest deeper legitimation strategies, one can interpret these arrangements in light of political and moral theory, arriving at a reconstruction of the normative assumptions the institution.132

that drive

they

Legitimation strategies do not prescribe a particular set of institutional arrangements, but and constrain possibilities. For example, as Richard Stewart has shown, the early models for the self-justification of the U.S. administrative state viewed discretion as the most pressing problem and designed institutions that would tightly constrain the exercise of such discretion by administrative officials.133 The ultimate inability of these models to constrain discretion spurred the development of an

create

(2008) (noting that accountability and legitimacy discussions are often intertwined, though they remain distinct concepts).

128. Grant & Keohane, supra note 115, at 30-31. 129. 130. 131.

In particular, see id. at 34-35. DAVID BEETHAM, THE LEGITIMATION OF POWER 37 (1991). Julia Black describes the manner in which institutions often generate their

own self-justificatory discourse. See generally Black, supra note 127.

132.

In addition to Beetham's work, the author has found helpful contributions by Jiurgen Habermas, Legitimation Problems in the Modern State, in COMMUNICATION AND THE EVOLUTION OF SOCIETY 178 (Thomas McCarthy trans., 1979); ROBERTO MANGABEIRA UNGER, LAW IN MODERN SOCIETY: TOWARD A CRITICISM OF SOCIAL THEORY 243-68 (1976); RODNEY S. BARKER, LEGITIMATING IDENTITIES: THE SELFPRESENTATION OF RULERS AND SUBJECTS IAN CLARK, LEGITIMACY IN INTERNATIONAL SOCIETY (2007); Robert Howse, The Legitimacy of the World Trade Organization, in THE LEGITIMACY OF INTERNATIONAL ORGANIZATIONS 355-407 (JeanMarc Coicaud & Veijo Heiskanen eds., 2001). This approach is developed more fully in a yet-to-be-published paper by the author.

(2003);

133.

Richard B. Stewart, The Reformation of American Administrative Law, 88

HARv. L. REV. 1667, 1670-75 (1975).

longer posed a fundamental

alternative strategy, the "interest-representation" model, which solved the problem of discretion by reconceptualizing institutions, threat.134 such that discretion no Discretion persisted, but it was rendered less problematic by a series of self- and court-imposed controls designed to subject power to miniature versions of the democratic process within agencies.135 Though it provided a solution to the problem of discretion, this approach is not necessarily superior, as it creates a new constellation of potentially insoluble problems in place of the old issues.136 It may be suggested this continuous cycle of problem-solution combinations better represents the nature of legitimation in modern governance teleological approach suggested by many authorities.137

than the

that

Through analysis of

the cluster approach's accountability mechanisms, one will glimpse a similar dynamic at work. The supervisory accountability structure built into the cluster approach represented a theoretically coherent attempt to ground the system's operation in the consent of members and the well-worn processes of the United Nations. These constitute a set of arguments that the system's power, in the form of institutional choice, is both well founded and properly controlled.

But, for reasons that will be discussed, this structure was set up to fail. In its place, one sees traces of a new apparatus based on horizontal accountability, which would co-opt many of the difficulties that faced the earlier structure, treating them as strengths. This solution, in turn, creates new problems. Inclusion of the affected population, which once stood as just one among many critiques facing humanitarianism, will emerge as the key variable upon which the legitimation of the system depends.

134.

See id. at 1712 (regarding this model, "decisions made after adequate consideration of all affected interests would have, in microcosm, legitimacy based on the same principle as legislation and therefore the fact that statutes cannot control agency discretion would become largely irrelevant").

Id. at 1676-88. Id.

135. 136. 137. On the latter, see generally THOMAS FRANCK, THE POWER OF LEGITIMACY (1990); JUTTA BRUNtE & STEPHEN J. TOOPE, LEGITIMACY AND AMONG NATIONS LEGALITY (2010); Allen Buchanan & Robert 0. Keohane, The Legitimacy of Global Governance Institutions, 20 ETHICS & INT'L AFF. 405 (2006).

INTERNATIONAL LAW: AN

INTERACTIONAL AcCOUNT

V. SUPERVISORY ACCOUNTABILITY IN THE UN HUMANITARIAN ARCHITECTURE

Institutional innovation and development in the humanitarian arena takes place in a tightly constrained normative landscape. Accountability strategies must navigate the preexisting normative demands on the institution, which often greatly constrain the types of mechanisms that may be imposed. This Part maps the constraints of autonomy and coordination on humanitarian institution building, before proceeding to outline the formal accountability structure of the cluster approach. As will be shown, the formal system was likely to fail not only because of logistical and practical constraints but also because it is fundamentally unable to cope with the values of coordination and autonomy that shape action in this field.

A. Autonomy and Coordination as Constraints on Accountability

The UN cluster approach depends for

its survival on the participation of a broad range of actors that possess no defined legal to work with the United Nations.138 Therefore, the obligation mechanism must be made to appear sufficiently attractive and justifiable (or legitimate) to secure the participation of the major humanitarian actors whom it purports to coordinate. This does not mean that the participants to be convinced. Donors and political actors, if convinced of the value of a centralized UN mechanism for humanitarian coordination, may be able to force reluctant NGOs to participate. At the same time, however, the Red Cross and other humanitarian NGOs wield significant normative influence that might be used to undermine any effort at coordinating relief activities.139 So, while the pathways for influence might be diffuse, the system must justify itself in order to work.140

themselves necessarily need

138.

See, e.g., U.N. Emergency Relief Coordinator and Under-Secretary-General

for Humanitarian Affairs, Inter-Agency Real-Time Evaluation of the Humanitarian Response to the Darfur Crisis 63 (Jan. 2006) (by Bernard Broughton et al.) ("Agencies cannot be forced to work within the parameters of a common plan-ultimately sector leads must persuade the majority of the value of a cohesive approach.").

139.

The receptivity of participating NGOs is, thus, a central concern of the cluster system and features prominently in the ongoing review process. The operations of the IASC represent a clear effort to bring the major stakeholders (i.e., NGOs) on board in designing this structure.

140.

justify

sufficient

It will not be

of humanitarianism, or even humanitarian coordination. The United Nations' stated purpose "to achieve international co-operation in solving international problems of a .. .humanitarian character" provides legal grounding for the organization's role in this context, but it does not legitimize any particular institutional arrangement. See U.N. Charter art. 1, para. 3. In the end, the institutional structure of the cluster approach must itself compose a normative argument as to why agencies ought to participate, in light of the relevant values at stake. In Professor Shaffer's words, "[The

the general enterprise

Any effort to impose accountability upon or within the cluster system will confront the problem of navigating between autonomy and effective coordination.141 This tension is particularly pronounced within the cluster system, where both concepts are closely tied to deeply held values and principles. On the one hand, effective coordination has emerged as the watchword of emergency response, and the system will face pressures to orchestrate the increasingly varied and numerous foreign and domestic actors engaged in major disasters.142 On the other hand, the system will also face significant ,pressure to preserve the autonomy of humanitarian actors, who continue to operate under competing sets of principles and compete for donors.143 In order to be perceived as normatively justifiable, any central effort to coordinate relief activities must hold at least the possibility for resolving the tension between these two impulses.

Though calls

for coordination have become

increasingly prevalent as the number of humanitarian actors has multiplied,144 this should not imply that actors are coalescing around a single correct approach to disaster response.145 By emphasizing the need for diversity and experimentation in approaches, humanitarian agencies have been relatively successful in delegitimizing any effort to wholly

pursuit of any substantive goal is necessarily mediated through different institutional processes that will affect outcomes, so that institutional analysis is required and such analysis must be comparative." Gregory Shaffer, Comparative Institutional Analysis and a New Legal Realism, 2013 WIs. L. REV. 607, 607-08 (reviewing the work of Neil Komesar).

Judge Katzmann has helpfully defined

terms of 141. organization theory. See Robert A. Katzmann, Note, Judicial Intervention and Organization Theory: Changing Bureaucratic Behavior and Policy, 89 YALE L.J. 513, 521-22 (1980).

this opposition

142.

See FISHER, supra note 4, at 150 ("Coordination is probably the most discussed issue in international disaster response. Yet, failures in this area remain a constant complaint both among international actors and between international actors and their domestic counterparts in affected states.").

See, e.g., ABBY STODDARD ET AL., OFFICE FOR THE COORDINATION OF HUMANITARIAN AFFAIRS, CLUSTER APPROACH EVALUATION 17 (2007) (noting problems with securing "the support of otherwise independent and operationally capable NGOs" for the cluster system); Mohammed Moshtari & Paulo Gongalves, Understanding the Drivers and Barriers of Coordination Among Humanitarian Organizations, 19-20, 25 (POMS 23d Annual Conf., Chicago, 2011), available at http://www.pomsmeetings.org/confproceedings/025/FullPapersfFullPaperfiles/025-0671.pdf (noting the possibilities for competition or coordination failures even among organizations that share the same principled outlook).

II., Apr. 20-23,

144.

Coordination now figures prominently in almost every evaluation and review of humanitarian responses. E.g., ALNAP, EVALUATING HUMANITARIAN ACTION USING THE OECD-DAC CRITERIA: AN ALNAP GUIDE FOR HUMANITARIAN AGENCIES § 3.6.1 (2006) ("[Cloordination is an important consideration in the evaluation of humanitarian action."); OECD DEVELOPMENT ASSISTANCE COMMITTEE, GUIDANCE FOR EVALUATING HUMANITARIAN ASSISTANCE IN COMPLEX EMERGENCIES 23 (1999).

145.

See supra notes 87-88 and accompanying text (explaining the divergence in

principle and practice among organizations).

integrate emergency response efforts under one hierarchical structure.146 In his review of "international disaster response law," David Fisher suggests that the failure of the International Relief Union in the early 1940s taught the international community to avoid "command and control" coordination mechanisms.147 Fisher argues that the resulting independence of the Red Cross and other humanitarian NGOs may be viewed as a "salutary effect" of the move away from centralization. 148

Humanitarian organizations and donors value diversity for different reasons. Aid actors have long been skeptical of the United Nations' emphasis on peacekeeping and peace building, an approach that is not always compatible with the fundamental humanitarian principle of neutrality with respect to antagonistic parties.149 Indeed, the cluster system in its early years suffered the criticism that the clusters were a UN-centric mechanism,150 raising concerns about the system's long term legitimacy and sustainability. The preferences of powerful donors such as states are also not neutral, as history indicates that they seek the ability to channel their money to multiple possible sources.151 Thus the cluster system must accommodate normative arguments for autonomy. Otherwise, actors might seek to undermine the coordination process by pointing to operational defects that harm affected populations or by emphasizing the value of diversity and experimentation.152

146.

See, e.g., Angelo Gnaediger, Director-General, Int'l Comm. of the Red Cross, Keynote Address, Int'l Conference of Voluntary Agencies, The Value of Diversity (Geneva, Switzerland, Feb. 1, 2006), available at http://www.icrc.org/eng/resources/documents/ statement/6m4lpw.htm ("In the face [o]f this enormous variety of humanitarian calls, the diversity of actors greatly enhances the flexibility and the appropriateness of the response."). 147.

FISHER, supra note 4, at 151. On the background of the International Relief Union, a treaty-based organization created under the auspices of the League of Nations, see P. MacAlister-Smith, Reflections on the Convention Establishing an International Relief Union of 1927, VOOR RECHTSGESCHIEDENIS [J. LEGAL HISTORY] 363 (1986) (tracing the establishment and collapse of the organization).

TIJDSCHRIFr

July

12,

148. 149.

FISHER, supra note 4, at 151. See Nicolas de Torrent6, Humanitarianism Sacrificed: Integration's False

Promise, 18 ETHICS & INT'L AFF. 3, 6 (2004) (arguing that "the hierarchy of priorities inherent in the coherence agenda often results in humanitarian interests being sacrificed or sidelined in the name of a 'greater good'); FIONA TERRY, CONDEMNED TO REPEAT? THE PARADOX OF HUMANITARIAN ACTION 24 (2002) ("Either aid is given without discrimination, or it is given in the interests of peace."). Addressing this principled critique has been a central focus of recent cluster reforms, and it has been a major driver in the push to invite independent NGOs to cochair operational clusters. NGOs & HUMANITARIAN REFORM PROJECT, THE PARTICIPATION OF NGOs IN CLUSTER Co-LEADERSHIP AT COUNTRY LEVEL: A REVIEW OF EXPERIENCE 3 (Feb. 2010).

150. 151. 152.

STODDARD ET AL., supra note 143, at 19. See sources cited supra note 147 and accompanying text. E.g., ICVA, DRC, supra note 56 (examining the rollout of the cluster

approach in Kinshasa in 2006).

B. Outlines of a Formal Supervisory Structure

an

elegant,

two-tiered

On paper, the cluster approach solves monitoring problems through structure hierarchical supervision.1s3 The in-country humanitarian coordinator, a UN official, appoints the cluster lead agencies and holds them responsible for ensuring effective coordination within their sectors.154 Though the cluster system does not alter the formal legal relationship between the United Nations and the relevant agency, the humanitarian coordinator could essentially fire a cluster lead by replacing it or embarrass the agency by releasing information about its activities. The humanitarian coordinator is, in turn, formally supervised by the head of OCHA in New York, who may hire and fire the in-country official.'55

The substance guiding this relationship is outlined in Part J.156 Lead agencies are responsible to the humanitarian coordinator largely for a number of general procedural and substantive considerations: inclusion of humanitarian actors, establishment of coordination mechanisms, coordination with local authorities, community participation, consideration of "cross-cutting issues," needs assessment, emergency preparedness, "planning and strategy development," application of humanitarian and human rights standards, monitoring resourcemobilization, training and capacity building, and provision of services as a "last resort."157 The substantive norms governing cooperation, crosscutting issues, and the like are expected to develop through repeated interactions within clusters at the global and local levels.'15 By producing or endorsing handbooks, toolkits, and guidelines, the IASC can exercise some control over the normative standards that inform cooperation.159 The humanitarian coordinator's responsibilities

reporting,

advocacy,

153.

The outlines of this structure are explained in IASC, CLUSTER APPROACH, supra note 20. On "supervisory accountability," see generally Grant & Keohane, supra note 115, at 36.

154. On the role of the humanitarian coordinator, see IASC, HC TOR, supra

note 27. 155.

The head of OCHA is also often referred to as the emergency relief coordinator. Strengthening the Coordination of Humanitarian Emergency Assistance of the United Nations, G.A. Res. 46/182, Annex, 34 U.N. Doc. A/RES/46/182 (Dec. 19, 1991).

See supra text accompanying notes 44-53 (describing the responsibilities of

.IASC, Cluster Lead TOR, supra note 44.

E.g., IASC, CLUSTER APPROACH, supra note 20, at 6 (noting that clusters

should implement standards set at the global level).

159.

E.g.,

INTER-AGENCY STANDING COMM., CIVIL-MILITARY GUIDELINES & REFERENCE FOR COMPLEX -EMERGENCIES (2008). These documents will often be formulated quite capaciously in order to obtain agreement, and it is not clear that they are regularly used. The minimal survey data compiled by the IASC shows mixed

156. cluster leads). 157. 158.

are phrased in similar procedural terms: articulating a "Common Humanitarian Action Plan," ensuring the coordination of clusters and the proper functioning of lead agencies, and establishing a mechanism for intercluster coordination. 160

This vertical structure replicates the legal basis for the cluster system itself.161 The authority of the IASC to create something like the cluster approach can be traced largely to a single resolution of the General Assembly in 1991.162 That resolution sketched a hierarchical system, which this accountability structure.63 The central innovation of the cluster system, established 15 years later, was the creation of lead agencies, which now occupy the ground-level tier of the accountability structure.

the backbone of

remains

implementation

The possibility of greater institutionalization of humanitarian activities may have been threatening to disaster-prone states, whose emergency authority might be threatened by a strengthened humanitarian response structure.164 This fear is observed in the practice of the cluster system, in which states have occasionally resisted sometimes successfully.165 Endeavoring to make the system directly responsive to a subordinate of the secretary-general, who is appointed by the General Assembly,166 suggests a desire to maintain some measure of state control over the system, rendering it more palatable to states. In addition, the hierarchical mechanism grounds the system in the internal law of the United Nations, as it is built on the General Assembly's powers to establish organs and offices.167 Also, the mere fact that the hierarchical structure mapped earlier forms of humanitarian organization within the United Nations may have been attractive: if the cluster innovation is to be seen as a technical

approach,

160.

IASC, HC TOR, supra note 27, at 2. The terms of reference, like other documents, set up certain broad substantive considerations: "age, gender, diversity, human rights, HIV/AIDS, and the environment." Id.

161.

Strengthening the Coordination of Humanitarian Emergency Assistance of

the United Nations, G.A. Res. 46/182, Annex, U.N. Doc. AIRES/46/182 (Dec. 19, 1991).

Id. Id. at Annex, 11 33-39.

162. 163. 164. Humanitarian institutions often function as an auxiliary to, functional substitute for, or challenge to state government in times of crisis. See generally PAUL HARVEY, TOWARDS GOOD HUMANITARIAN GOVERNMENT: THE ROLE OF THE AFFECTED STATE IN DISASTER RESPONSE (Humanitarian Pol'y Grp., Pol'y Brief No. 29, Sept. 2009).

165.

Samir Elhawary & Gerardo Castillo, The Role of the Affected State: A Case Study on the Peruvian Earthquake Response 12 (Humanitarian Pol'y Grp., Working Paper, 2008) (citing political opposition as a main reason why the cluster approach was not activated in the 2007 Peruvian earthquake).

166. 167.

U.N. Charter art. 97. U.N. Charter art. 22.

AIANAGING THE "REPUBLIC OF NGOS'

improvement, rather than a transfer of leadership power from the host states to international and nongovernmental organizations, it appears natural to subject the system to familiar procedures, which are responsive to a familiar set of interests.

C. The Breakdown of the Formal Structure

The weakness of the formal structure is among the most widely failings of the cluster approach.168 The oversight

recognized mechanism is of course vulnerable to a range of logistical problemsOCHA, which exercises the top level of supervisory responsibility, cannot realistically be expected to "line manage" the large number of coordinators around the globe.'69 But this subpart suggests that the deeper reasons for the system's failure are normative. The vertical accountability structure reflects a view of humanitarian practice that is incommensurable with the strong commitment to autonomy that is shared, for different reasons, by humanitarian practitioners and donors. The mechanism's inability to resolve the tension between coordination and autonomy opens the way for new institutional solutions, which are the subject of the following Part.

So far, this Article has discussed the humanitarian coordinator as if the coordinator is a discrete official who is responsive to the head of OCHA in New York, but this is generally not the case. Often, the coordinator also serves as the UN "Resident Coordinator," who is responsible for overseeing development operations.170 This position comes with a parallel chain of command, with the resident team and ultimately through a regional coordinator reporting responsive to the UN Development Group, which involves a different set of actors.171 The coordinator may play other roles, with additional supervisors, in the context of peacekeeping missions.172 Critics argue to the subordination of inevitably lead that these other roles

168.

See NGOs & HUMANITARIAN REFORM PROJECT, SYNTHESIS REPORT: REVIEW OF THE ENGAGEMENT OF NGOs WITH THE HUMANITARIAN REFORM PROCESS 3 (2009) ("The UN has continued to appoint unqualified Humanitarian Coordinators (HCs) who do not adequately understand humanitarian action; who underestimate the importance of NGOs; who do not understand the critical importance of partnership ... ."); PAUL HARVEY ET AL., ACTIVE LEARNING NETWORK FOR ACCOUNTABILITY AND PERFORMANCE IN HUMANITARIAN ACTION, THE STATE OF THE HUMANITARIAN SYSTEM: ASSESSING PERFORMANCE AND PROGRESS-A PILOT STUDY 49 (2010) (noting a lack of leadership).

169.

Office for the Coordination of Humanitarian Affairs [OCHA], Strengthening

the HC System: The Unfinished Agenda 2 (paper presented to the IASC Working Grp., Mar. 2009). 170.

Strengthening the Coordination of Humanitarian Emergency Assistance of the United Nations, G.A. Res. 46/182, Annex, T 39, U.N. Doc. A/RES/46/182 (Dec. 19, 1991) ("[The resident coordinator should normally coordinate the humanitarian assistance of the United Nations system at the country level.").

171. 172.

OCHA, supra note 169, at 2-3. Id. at 3-4.

humanitarian or developmental motives.173 -Despite several proposals to combat these problems,174 this dual structure is likely to persist.75

security, military,

political,

concerns

By

splitting

the coordinator's

time, or office, between humanitarian, developmental, political, and potentially peacekeeping roles, the humanitarian coordinator personifies the range of values at the heart of UN activities, which humanitarian organizations do not necessarily share. Thus, as one moves up the chain of accountability, the humanitarian mission becomes mixed with other goals that may be seen to undermine humanitarian principles.176 In the end, these other functions are not severable; they are central to the identity of international problem solver. The the United Nations as an centralized UN structure will always be torn between the project of securing peace, fostering a state's economic development, and impartial humanitarian assistance. The providing neutral and humanitarian coordinator will never be entirely free from this conflict and, therefore, will often be kept at arm's length by the non-UN humanitarian organizations. Coherence with the broader range of UN activities thus constitutes both a benefit and a curse for the cluster is at the same time an essential feature of the approach-it

173.

E.g., Vanessa Humphries,

Improving Humanitarian Coordination:

Common Challenges and Lessons Learned from the Cluster Approach, J. HUMANITARIAN ASSISTANCE, Apr. 30, 2013, available at http://sites.tufts.edul jha/archives/1976 (noting threats to independence and impartiality arising from close ties between NGOs and UN agencies).

174.

them

training

For example, a pool of potential humanitarian coordinators has been developed, which identifies possible candidates and gives in humanitarian affairs. Humanitarian Coordination Pool, HUMANITARIAN RESPONSE, https://www.humanitarianresponse.infolhumanitarian-coordination-pool (last visited Dec. 27, 2013) ("The IASC HC Pool, established in July 2009, is a roster of high caliber humanitarian professionals from UN agencies, the Red Cross/Red Crescent movement, IOM and NGOs who have been screened by the IASC as potential candidates for humanitarian coordination leadership positions."). This has generated greater calls for transparency in the selection of pool members, for greater NGO input, and for the identification of more pool members from outside the UN system. This will only have substantial impact, however, to the extent the humanitarian role is separated from the development that figures prominently in many reform efforts. E.g., SAVE THE CHILDREN, AT A CROSSROADS: HUMANITARIANISM FOR THE NEXT DECADE (2010). Another solution is to appoint deputy humanitarian coordinators from the aid profession. But, where this is used to make up for a lack of humanitarian knowledge at the top level, this solution is highly dependent on the creation of a strong bond between the deputy and the humanitarian or resident coordinator; if the top-level official is uninterested in humanitarian issues or principles, then the deputy creates an organizational way to marginalize these problems.

a separate person, a measure

role and instilled

176.

See, e.g., TERRY, supra note 149, at 23-26 (detailing tensions between the

demands of peace and the "humanitarian imperative").

coordinating mechanism and a grave threat to the autonomy, neutrality, and independence of humanitarian NGOs.177

itself

terms,

In practical

this conflict manifests

the observation that humanitarian coordinators often lack the interest, expertise, and information to effectively oversee the operations of clusters.178 This lack of oversight may undermine the effectiveness of the cluster approach, but it does not neutralize its power. To the contrary, because the power of the system arises in large measure from its appointment of leaders and distribution of competences,179 the effects of institutional choice may actually be magnified when the appointed leaders are not subject to the prescribed supervision. The types of policies implemented, and the groups of actors who have access to inner policymaking circles, will be that much more likely to be determined by the individual characteristics of the leader and organization in charge of a particular sector.

Actors critical of the failure of the formal accountability structure are joined by another group of advocates who work from the perspective of grassroots organizing and affected populations. Such groups have long been critical of the cluster approach for shutting out local groups and affected persons, arguing, for example, that victims' rights groups had been excluded from UN facilities in Haiti during cluster meetings, that these meetings were conducted in languages spoken only by foreign workers, and that the perspective of local organizations was continually squashed.80

This is not a new critique: the failures of the 1990s led many to question whether the culture of the "humanitarian international" to successful disaster response.'8' And constituted an obstacle grassroots advocates are likely uninterested in seeing the cluster system succeed in its current form because the vertical accountability structure points away affected by

the persons most

from

177.

On the normative value of independence, see Daniel Thirer, Dunant's Pyramid: Thoughts on the "Humanitarian Space," 89 INT'L REV. RED CROSS 47, 58 (2007). See also de Torrent6, supra note 149, at 3-5 (arguing that integration threatens typical humanitarian principles).

178.

STEETS ET AL., supra note 10, at 45; see also KAUFFMANN & KRUGER, supra note 42, at 38-39 (noting the weak leadership from OCHA, arising from a gap in staffing). 179. 180.

See supra Part III. E.g., Melinda Miles, Assumptions and Exclusion: Coordination Failures During the Emergency Phase, in TECTONIC SHIFIS: HAITI SINCE THE EARTHQUAKE 45 (Mark Schuller & Pablo Morales eds., 2012) (arguing that Haitians were excluded from the cluster system's discussions); Maura R. O'Connor, Does International Aid Keep Haiti Poor?, SLATE (Jan. 7, 2011), http://www.slate.com/articles/news-and-politics/ dispatches/features/2011/does internationalaidkeep-haiti poor/the un cluster syste m is asbad as itsounds.html (describing the cluster system as "incomprehensible and dysfunctional"); Sontag, supra note 2, at Al.

humanitarian action-the victims-toward international UN officials and toward New York. Nonetheless, the failure of the formal accountability system creates the space for reimagining institutional structure, in which advocates of greater victim participation may be poised to play a crucial role.

VI. THE PROMISES AND CHALLENGES OF PEER REVIEW IN THE CLUSTER SYSTEM

Instead, clusters

Looking for new modes of accountability in the cluster approach, recent reviews have embraced the opportunities that it creates for peer review.182 The most recent evaluation states: "In all case study countries bar one, accountability to the Humanitarian Coordinator is minimal. have to make valuable contributions to strengthening peer accountability."83 The modes that these peer review processes take appear to be quite varied, ranging from informal lessons learned and reviews of funding proposals, to more structured interactions where standards and recommendations are developed and then forwarded to the next highest level in the structure.

started

In general, what consultants are calling peer review is really a hodgepodge of procedures, most of which amount to little more than peer pressure. The role of network-style interactions in encouraging socialization, argumentation, and harmonization of policy is wellknown,'84 and many of these theories may explain some successes of the cluster approach.85 But, as beneficial as these dynamics are, it is difficult to imagine that they would substitute for a regularized, dependable oversight structure.186

The real promise of peer review lies in its potential to foster a more routine decentralized system of peer monitoring, benchmarking, and reflexive revision of goals. This new architecture, it might be hoped, would provide a for hierarchical

functional substitute

182. 183. 184.

STEETS ET AL., supra note 10, at 69, 87. Id. at 44. See generally SLAUGHTER, supra note 64, at 195-212 (discussing network-

style interactions in international organizations).

185. 186.

for pressing

See discussion supra notes 62-69 and accompanying text. I

thank Richard B. Stewart

this point. As Georgios Dimitropoulos points out, however, peer pressure and trust relationships make a positive contribution to regulation under certain conditions. Georgios Dimitropoulos, Peer Reviews Between Institutions, Note, Workshop on Analyzing and Shaping InterInstitutional Relations in Global Governance at N.Y.U. (Apr. 16, 2012) (on file with author); see also SLAUGHTER, supra note 64, at 198-200 (explaining the benefits of socialization). The important role for law, then, is to ensure that the development of such bonds is not left entirely to individual psychology, charisma, and chance. See generally Rebecca M. Bratspies, Regulatory Trust, 51 ARIZ. L. REV. 575 (2009) (examining the relationship between law and regulatory trust).

arising

autonomy

the demands of supervision, while adhering more closely the humanitarian independence profession.187 Peer review would work by mitigating the effects of harmonization and institutional choice; by subjecting the policies developed within clusters to regular, ongoing review; and by empowering cluster members with information about successful and unsuccessful strategies elsewhere, it is hoped that the problems of weak leadership and bad policy can be corrected through argument and the UN apparatus sought its legitimacy through a hierarchy that could theoretically be responsive to the complaints of states and other interests; in the reimagined system, the United Nations emerges as a convener and orchestrator of a problem-solving enterprise.189 This may relax some of the demands directed toward the top of the hierarchy and shift the focus to direct participation and deliberation at lower levels.190

innovation.88 Whereas,

in the formal structure,

187.

See GrAinne de Bdirca, Robert 0. Keohane & Charles F. Sabel, New Modes of Pluralist Global Governance, 45 N.Y.U. J. INT'L L. & POL. 723, 786 (2013) ("Experimentalist Governance makes possible a form of forward-looking or dynamic accountability unavailable in traditional, principal-agent regimes .... .").

188.

See Joshua Cohen & Charles F. Sabel, Global Democracy?, 37 N.Y.U. J. INT'L L. & POL. 763, 779 (2005) (In such a system, "decisionmaking works through mutual reason giving. Deliberation subjects the exercise of collective power to reason's discipline, to what Habermas famously described as 'the force of the better argument,' not the advantage of the better situated.").

189.

See generally Kenneth W. Abbott & Duncan Snidal, International Regulation Without International Government: Improving 10 Performance Through Orchestration, 5 REV. INT'L ORGS. 315, 325-26 (2010); Kenneth W. Abbott & Duncan Snidal, Strengthening International Regulation Through Transnational New Governance: Overcoming the Orchestration Deficit, 42 VAND. J. TRANSNAT'L L. 501 (2009) (arguing that the orchestration of nonstate actors is less challenging than the orchestration of states, even absent hierarchical enforcement power).

190.

As noted in Part IV, the scope of accountability is contested. There is no question that the employee-employer type of relationship explored in Part V constitutes a type of accountability mechanism under any definition. But whether the horizontal form of control discussed here qualifies is subject to debate. The peer review process described below functions on the incorporation of diverse actors into the regulatory process itself, in order to foster debate and self-correction by the network. Because the mechanism relies on the inclusion of voices from actors who otherwise are heavily dependent on the services of cluster members, it may not qualify, according to some definitions, as an accountability mechanism. See Enrique Peruzzotti, Civil Society, Representation, and Accountability: Restating Current Debates on the Representativeness and Accountability of Civic Associations, in NGO AccouNTABILITY, supra note 113, at 41, 54 ("Not only are the basic conditions of an accountability relationship not present (exchange among two actors one of whom holds rights of superior authority, autonomy of the account holder, and so on), but ... the targets of [NGO] intervention not only lack equal standing, but too often stand in a relationship of extreme dependency with regard to the material goods or services that the organization provides."). This dependency relationship is highly relevant to the discussion that follows, but the debate over terminology would affect none of the conclusions here.

/VoL. 47239

This Part first introduces the practical obstacles of peer review in order to emphasize the extensive normative transformation that would be required to make such a structure work. Second, this Part draws on theories of experimentalism in regulatory governance to sketch an alternative-not nearly realized in practice-that would effect such a transformation. A final subpart reflects on the emergence of victim participation as the central legitimation problem in this new humanitarian architecture.

A. Peer Review and the Autonomy-Coordination Tension

The idea that, freed from the tethers of vertical supervision, humanitarians might take it upon themselves to control for the effects of the cluster system may not be very comforting. The development of robust peer review in the clusters has been hampered in part by a noted resistance among cluster participants to police each other's activities.191 Although rare cases show agencies setting targets and holding each other to account in meeting them,192 the clusters are generally not seen as a forum for mutual monitoring.'93 Michael Dorf and Charles Sabel have noted that mutual-monitoring structures are dependent on the "willingness of all participants to

at

191.

42-43, available at

STEETS ET AL., supra note 10, at 52. In recent years, humanitarian groups have begun to overcome this problem through a novel peer review process developed outside the clusters. See About the Inter-Agency Standing Committee: Steering Committee for Humanitarian Response (SCHR), INTER-AGENCY STANDING COMM., http://www.humanitarianinfo.org/iasc/pageloader.aspx?page=content-about-schr (last visited Dec. 28, 2013) (explaining this practice); Eva Von Oelreich & Yoma Winder, The SCHR Peer Review Process: Oxfam's Experience, HUMANITARIAN EXCH. MAG., Apr. 2006, magazine/issue-33/the-schr-peer-review-process-oxfams-experience. The method was first developed by the OECD and is applied in studies that focus on specific topics, such as sexual abuse or accountability to affected populations. Under this method, agencies work in groups of three, where each agency is reviewed by the other two in its group, guaranteeing a kind of reciprocity. STEERING COMM. FOR HUMANITARIAN RESPONSE, SCHR PEER REVIEW ON ACCOUNTABILITY TO DISASTER-AFFECTED POPULATIONS: AN OVERVIEW OF LESSONS LEARNED 20-21 (2010). The group also hires an outside consultant to facilitate the process and prevent collusive behavior. Id. The review results in a set of reports and "lessons learned," which remain private, as well as a synthesis paper for external audiences, which is largely sanitized of any direct reference to a specific failure by any NGO.

http://www.odihpn.org/humanitarian-exchange-

This process demonstrates what might be demanded of any kind of peer review conducted under the cluster system. But it should be noted that these reviews are timeconsuming, and only three have been initiated since the method was created in 2002. See IASC, CLUSTER APPROACH, supra note 20. Moreover, this is a process controlled entirely by the Steering Committee for Humanitarian Response, a group of major NGOs. It does not include UN actors or the IOM. Because of a strong adherence to the principle of independence, it is unlikely that the participating agencies would have agreed to a similar process facilitated by UN agencies.

192.

KROGER & STEETS, supra note 94, at 26 n.44 (noting that this took place in

the response to a 2009 drought in Gaza).

193.

STEETS ET AL., supra note 10, at 52.

disclose information in view of the investigations of others."194 One imagine several reasons, some principled and some selfcan interested, why humanitarian organizations may be unlikely to develop such an attitude.

also

refer

state or

fourth pillar

incorporated as a

the United Nations-they

The foregoing Parts have already noted the importance of independence and autonomy to the humanitarian enterprise. These principles not only instill a desire to be free from the yoke of any powerful to humanitarian institutions' independence from each other. In 2007, concerned that the cluster approach was being dominated by UN concerns for integration, NGOs pressed for a set of Principles of Partnership, which were to humanitarian reform.195 While the principles do not foreclose the possibility of critique, the independence of all partners in the humanitarian system suggests that there would be some resistance to the idea of recasting the cluster approach as a mode of continued monitoring and reporting.196 ideological backing or justification to actors that wish to circle the wagons for more self-interested reasons.'97 The realities of humanitarian action may provide a rational temptation among aid actors to engage in cartel-like behavior with respect to negative information. Given the desperate situations created by conflicts and environmental disasters, it is "always insufficient"198 and perhaps often harmful.199 One can also assume that, because donors expect humanitarian actors to "do good," open

The principle of independence may also provide

that humanitarian action

their expressed desire

to maintain

to assume

is safe

194. Michael C. Dorf & Charles F. Sabel, A Constitution of Democratic

Experimentalism, 98 COLUM. L. REV. 267, 338 (1998).

195.

"Equality requires mutual respect between members of the partnership irrespective of size and power. The participants must respect each other's mandates, obligations and independence and recognize each other's constraints and commitments. Mutual respect must not preclude organizations from engaging in constructive dissent." Press Release, Global Humanitarian Platform, supra note 67.

SYSTEM 66 (2012) (noting "agency resistance" to peer review processes).

197.

In other words, one comes to expect problems in the peer review process regardless of whether one starts from a principled or constructivist orientation or a rationalist one. Cf. Tim BUthe, Solomon Major & Andr6 de Mello e Souza, The Politics of Private Foreign Aid: Humanitarian Principles, Economic Development Objectives, and Organizational Interests in NCO Private Aid Allocation, 66 INT'L ORG. 571 (2012) (arguing that organizational interests do not constitute a major driver of NGO aid allocation, which is better explained by humanitarian principle and constructivist theory).

(quoting an ICRC official).

199.

See TERRY, supra note 149, at 25 (criticizing the zero-sum nature of the "do no harm" approach to aid); Alex de Waal, The Humanitarians' Tragedy: Escapable and Inescapable Cruelties, 34 DISASTERS S130 (2010) (noting the triage-like quality of humanitarian action).

VANDERBILTOURNAL OF TRANSNATIONAL LAW

and honest reporting of failures is likely to harm an organization's ability to raise funds.200 One might further assume that any agency operating in the field would possess some information about failures by its peers and that any time Agency Y suffers criticism from NGO X, it would be equipped to respond in kind. Such a vicious cycle could be expected to create an overall negative impact on the amount of money going to the humanitarian enterprise. The best way to avoid this cycle is to generally avoid singling out peer agencies and to avoid referencing of performance.201 Though one does see defections, the practice of IASC humanitarian this assumption of cartel behavior.202

reviews seems generally consistent with

organizations

reviews

specific

critical

The alignment of principle with self-interest suggests the development of a culture that may be generally unwilling to engage in searching self-criticism. Alex de Waal put his critique of UN humanitarian activities in particularly forceful terms:

or

contain

insights

powerful

[self-evaluations]

Some strong recommendations, but there is no mechanism for enforcing 'learning the lessons'. In fact, critical evaluations are used for the opposite purpose: they can be brought out later to defuse new criticisms with the riposte that the critic is not saying anything new. Repetition is a constant difficulty faced by critics of the UN specialized agencies: a critique repeated many times may be valid but is readily ignored because it has become boring. As well as concealed errors the agencies have (rarer) secret successes, but lack of accountability means that successful innovations are only occasionally recognized (and rarely replicated).203

NGOs, de Waal argued, suffered similar accountability deficits and learning disabilities.204

What is needed is a more robust theory of peer review that dissolves the tension between autonomy and coordination at both the practical and the normative levels. If mutual monitoring is to control the effects of institutional choice, and therefore provide a functional substitute for more traditional modes of accountability, it must create learning and revision of institutional realistic possibilities for defective policies. As Larry Minear, a leading researcher in the field

200.

See TERRY, supra note 149, at 235-37 (exploring the economics of humanitarian action and arguing that the negative side effects of humanitarian action are exacerbated by organizational culture and the "culture of justification that permeates the entire aid community"). This does not necessarily mean that NGOs will be dishonest, only that there is some incentive to be.

201.

Id.; see also HARLOW & RAWLINGS, supra note 116, at 307

("With participants rendered complicit in decisions, there is a risk of denegration into a complacent 'old boy network'-the accountability function blunted by mutual interestand there are obvious problems of transparency.").

202.

The three-month review of humanitarian activities in Haiti, for example, refers to problems in various clusters but judiciously avoids singling out agencies. IASC, Response to the Humanitarian Crisis in Haiti, supra note 57, at 16-29.

203. 204.

DE WAAL, supra note 181, at 72. Id. at 80-81.

of humanitarian policy, put institutional change."205

it:

"[S]erious

learning

requires

B. Experimental Humanitarian Institutions: Sketching an Alternative Model

that seek alternatives both

term experimentalism implies a range of regulatory techniques to command-and-control regulation and to the minimalism of deregulatory approaches.206 Experimentalist strategies are based on a set of management that might be expressed as subsidiarity, principles inclusion, review.207 These strategies grant broad reflexivity, and peer discretion to local-level actors to pursue certain goals, with very little steering from the top down.208 The local units are meant to ensure the broadest possible participation by stakeholders, both inside and outside of the public administration apparatus.209 All of the norms generated through this process should be subject to periodic revision; this includes the specific practices of local units, as well as the means for measuring performance, the decision-making procedures, and the overarching goals.210

Some form of peer review is necessary to get this process going, but it is not sufficient.211 The price of broad delegation and discretion to innovate is constant reporting and monitoring.212 Information is

205.

Larry Minear, Learning to Learn, OCHA Seminar on Lessons Learned on Humanitarian Coordination, Apr. 3-4, 1998, at 9 (quoted in TERRY, supra, note 149, at 232).

206.

For the origins of the concept, see Dorf & Sabel, supra note 194. Experimentalism is often considered as part of an array of so-called new governance techniques. See de Bdrca, supra note 22, at 228 (distinguishing experimentalism from the broader concept of new governance).

207.

This list is the author's, but it draws directly from the description in Sabel

& Simon, supra note 22, at 79.

208.

See Dorf & Sabel, supra note 194, at 322 ("Above all, an experimentalist

regime gives locales substantial latitude in defining problems for themselves.").

209. 210.

Sabel & Simon, supra note 22, at 79; de Bdrca, supra note 22, at 228. See JEAN L. COHEN, REGULATING INTIMACY: A NEw LEGAL PARADIGM 15179 (2002) (arguing that reflexive law demands compliance with a system of principles without demanding specific outcomes); Michael C. Dorf, The Domain of Reflexive Law, 103 COLUM. L. REV. 384, 398-400 (2003) (arguing, in response to Cohen, that reflexive regulation can take place even without agreement on goals and that "rolling regulation" may lead to "the transformation of ends" in the long run); William H. Simon, Toyota Jurisprudence: Legal Theory and Rolling Rule Regimes 13 (Columbia Pub. L. & Lgl. Theory Working Paper Grp., No. 04-79, 2007) ("The phrase kaizen, or continuous improvement, connotes that process be revised in the course of its execution."). 211.

E.g., de Birca, Keohane & Sabel, supra note 187, at 786 (noting the

importance of peer review).

212.

See Dorf & Sabel, supra note 194, at 288 ("[The price communities must and should want to pay ... for the right to experiment is to provide individuals in their own and other jurisdictions with information to judge their performance ....").

learn

so that

local groupings can

innovations.213 This process contributes

from each other's pooled experiments and to comparative assessments across jurisdictions or problem areas and to continued debate within local units as to whether a competing approach is superior and should be adopted.214 The process thus depends on a blurring of boundaries, where actors simultaneously take on the role of regulator, monitor, and regulated entity.215 The result is a relatively concentrated form of peer review, where procedures for mutual evaluation among actors are seen as central to the success of the enterprise.216

autonomy-coordination

In theory, the experimentalist approach promises to dissolve, or tension. The least dampen, at experimentalist model depends on the fact that the members of the system will be pursuing their own autonomously generated programs and policies, rather than implementing a set of performance standards imposed from the top down. For this reason, experimentalism is attractive in places where top-down regulation is seen to have failed (e.g., primary education in the United States) and in arenas where the independence and autonomy of actors is jealously guarded (e.g., the European Union).217 Likewise, an regulatory policy experimentalist humanitarian architecture would rely on the independence of NGOs, rather than resist it.

In addition to the general alignment of values, readers may have already recognized certain design aspects of the cluster system that resemble the four dimensions of experimentalism-subsidiarity, inclusion, peer review, and reflexivity-described above. In terms of subsidiarity, much of the clusters' successes have been seen at subnational levels, where provincial or municipal clusters are able to adapt to local conditions and demands.218 Broad guidance from the global or national level would leave significant room for innovation, assuming of course that the participating international agencies are also given sufficient flexibility by their headquarters or regional

213.

See Sabel & Simon, supra note 22, at 79-80 ("The center provides services and inducements that facilitates this disciplined comparison of local performances and mutual learning among local units.").

214.

See Dorf & Sabel, supra note 194, at 300-01 (describing the use of

information pooling and comparison within and among firms).

215.

See Orly Lobel, The Renew Deal: The Fall of Regulation and the Rise of Governance in Contemporary Legal Thought, 89 MiNN. L. REV. 342, 391 (2004) (noting the interactions among legislation, implementation, enforcement, and adjudication); see also Jason M. Solomon, New Governance, Preemptive Self-Regulation and the Blurring of Boundaries in Regulatory Theory and Practice, 2010 WIS. L. REV. 591 (analyzing "new governance" as a process of blurring lines).

216.

On the related concept of peer accountability, see Grant & Keohane, supra

note 115, at 37.

217.

See Joshua Cohen & Charles Sabel, Directly-Deliberative Polyarchy, 3 EUR. L.J. 313, 315-16 (1997) (suggesting areas where the experimentalist model has been successful). 218.

Thanks to Paul Christian Namphy for pressing this point.

directors. In principle, the cluster approach is also designed to be broadly inclusive, directing lead agencies to ensure the inclusion of all key humanitarian partners.219

Where successful, the cluster approach has also triggered peer review and reflexive revision of humanitarian policy. The Kivu regions of the Democratic Republic of the Congo (DRC), which rank among the most dangerous, nonetheless provided a success story for reflexive innovation.220 Provincial clusters developed a systematic practice of issuing recommendations to each other, which were then followed up in an intercluster meeting.22' It was noted that this process was bolstered by the relatively substantial amount of funds directed toward eastern DRC, in comparison to other regions, and the resulting capacity of OCHA to act as a coordinator.222

But robust peer review is frustrated by the limited ability of the system to pool and transfer information. While peer review succeeded in the Kivu region of the DRC, the overall national response was largely unable to benefit from these insights because of a lack of strategic coordination at the country level in Kinshasa.223 In areas such as the DRC, where transportation and communications prove difficult, the capacity of OCHA to move and store information across regions is weakened.224 The flow of information among policymaking this perspective, as continuous crucial sites becomes to evaluate and revise argumentation, and monitoring serves programs, to another and generally catalyzing the experimental process. The more OCHA can act as an information gatherer and pooler, the more effective an experimental structure can be and the less reporting requirements will interfere with the daily work of disaster response.225

lessons from one site

transmitting

from

219.

IASC, Cluster Lead TOR, supra note 44, at 1. Here, this Article glosses over what is a fraught and controversial issue. The deep-seated problems with inclusion in the cluster approach are addressed in subpart C.

220.

See ANDREA BINDER, VPRONIQUE DE GEOFFROY & BONAVENTURE SOKPOH, IASC CLUSTER APPROACH EVALUATION, 2ND PHASE: COUNTRY STUDY-DEMOCRATIC REPUBLIC OF CONGO 31-34, 38-39 (2010).

221. 222.

Id. at 31. Id. at 32. A strong OCHA office in the nearby city of Goma reduced access

problems. Id. at 39.

223.

Id. at 32-33 ("[here was no strategic support or guidance from the

national level on how to address this sensitive coordination challenge.").

224. 225.

Id. at 25. OCHA itself is not an operational agency, in the sense that it does not provide services, and, therefore, it is intuitively well suited to this background role. Note also that the information-gathering role of OCHA in this context would not the the current burden necessarily be equivalent humanitarian coordinator's oversight over cluster leads. The role of OCHA would be less focused on informing the humanitarian coordinator of what has gone on in recent cluster meetings and more on making information about groups' various activities more widely available to other participants.

it carries of supporting

VANDERBIL TJOURMAL OF TRANSNA TIONAL LAW

In addition to dramatically increasing the informational capacity of OCHA, a successful experimental structure should attempt reforms to the cluster approach along at least four lines. Not all of these would require explicit changes in policy-an experimental system, like other forms of late capitalist regulation, is as much an attitude as a set of rules.226 But codifying the following considerations into the next revisions of IASC guidance on the cluster system may allow attitudes to follow policy. Note also that the following directions can be phrased in only general terms; a governance framework that emphasizes context sensitivity and adaptation should not devolve into a prescriptive, off-the-shelf model for institutional design.227 In short, the system must develop for making norm generation routine, performance monitoring, "rolling rulemaking" across geographic scales, and a rethinking of the United Nations' supervisory role.228

clear procedures

1. Active and Clear Rulemaking

If the cluster approach is to embrace an experimentalist model as an alternative to vertical hierarchy, the national and local cluster meetings must be used for more than information sharing.229 Active rulemaking by local sites is critical to unlocking the diagnostic potential of experimental governance, as the introduction of new rules allows actors to test hypotheses and plans.230 Moreover,

Cf.

226.

IAN AYERS & JOHN BRAITHWAITE, RESPONSIVE REGULATION: TRANSCENDING THE DEREGULATION DEBATE 5 (1992) (defining responsive regulation as an "attitude").

227.

See de Biirca, supra note 22, at 236-38 (discussing the critiques of new

governance methods); Cristie Ford, New Governance in the Teeth of Human Frailty: Lessons from Financial Regulation, 2010 Wis. L. REV. 441, 484 (arguing that "new governance methods may simply not be feasible in some contexts"). For a critique of experimentalism along these lines, see generally William E. Scheuerman, Democratic Experimentalism or Capitalist Synchronization? Critical Reflections on DirectlyDeliberative Polyarchy, 17 CAN. J. L. & JURISPRUDENCE 101 (2004).

228.

Cf. Javier Barnes, Towards a Third Generation of Administrative Procedure, in COMPARATIVE ADMINISTRATIVE LAW, supra note 121, at 336, 342 ("The need for procedural rules is in direct proportion to the lack of substantive provisions."); Hari M. Osofsky, Multidimensional Governance and the BP Deepwater Horizon Oil Spill, 63 FLA. L. REV. 1077, 1137 (2011) (emphasizing the need for regulatory responsiveness and inclusion across scales).

229.

See STEETS ET AL., supra note 10, at 34 (stating that meetings spend too much time catering to the information needs of agencies that have weak field capacity and that the discussions are often too abstract and do not disseminate helpful guidance or information); KAUFFMANN & KRGER, supra note 42, at 35 (noting that many intercluster meetings fail to move past information sharing); see also KROGER & STEETS, supra note 94, at 24 ("Regular inter-cluster meetings take place in Gaza and Jerusalem, but these are not seen as very useful by most humanitarian actors because they do not systematically focus on inter-cluster gaps or inter-disciplinary issues and do not focus on joint activities or programming.").

230.

See Dorf & Sabel, supra note 194, at 350-54 (discussing "rolling best-

practice rules").

idea originates

solutions.231 This

if they are combined with

experimental rules, counterintuitively, should be as precise as possible, in order to facilitate the diagnosis of problem areas and the identification the manufacturing sector, where the introduction of new, highly precise standards allows- companies to measure performance and to learn under controlled conditions,232 but the concept has seen some success in service delivery as well.233 The extremely capacious standards developed by some clusters do not necessarily facilitate error detection and problem solving, because it will not be clear whether all actors are faithfully reproducing the same experiment. Nonetheless, even broad standards may be consistent with an experimental approach reporting and monitoring.234 The

and hierarchical rulemaking lie in the type of process and obligation associated with the rule. First, as mentioned above, such rules are generated locally, as the result of a collaborative process among many actors.235 Second, experimental rules are "indicative or presumptive rather than mandatory."236 The idea is not to create rules for service provision that are followed rigorously by all cluster participants, even at the local level. Rather, the rules create a point of departure for "Strength/weakness" further analysis should become a first step toward developing new action plans and programs that depart from older, inappropriate standards and practices.237 Thus, rules for setting up a governance structure for refugee camps can be altered when it becomes clear that the agreedreinforce gender or power disparities. upon standards would

two critical differences between

experimentation

experimental

innovation.

frequent

231.

See Simon, supra note 210, at 16 ("[N]orms are always as articulated as possible, but they are not applied consciously in a way that would frustrate their purposes."). 232.

See id. at 17 ("The duty to articulate forces the actors to reflect on what they are proposing to do and to communicate it as precisely as possible to their peers."). See Sabel & Simon, supra note 22, at 91-92 (describing successes in child

233.

service provision).

234.

Some areas of governance in Europe have developed an approach known as comply or explain, in which regulation serves as a safe harbor, compliance with which excuses the regulated entities' duty to report on their conduct. E.g., U.K. FINANCIAL REPORTING COUNCIL, WHAT CONSTITUTES AN EXPLANATION UNDER 'COMPLY OR EXPLAIN'? REPORT OF DISCUSSIONS BETWEEN COMPANIES AND INVESTORS (Feb. 2012). This works where the rules are such that compliance can be easily gauged and, thus, favors precision in drafting. If a system relies heavily on general standards, then gauging compliance will be more difficult, and experimentation may best be facilitated by a simple duty to explain. But see SPHERE HANDBOOK, supra note 81, at 8-9 (suggesting a principle of comply or explain for humanitarian indicators).

235. 236. 237.

See supra text accompanying notes 207-25. Sabel & Simon, supra note 22, at 80. See BINDER & GRONEWALD, supra note 42, at 31 (noting the usefulness of "strength/weakness" analysis for improving accountability of lead agencies to cluster members, but adding that accountability was generally weak).

VANDERBIL TJOURNAL OF TRANSNA TIONAL LAW

Crucially, however, these innovations should be coupled with a practice of reporting back on the problems identified and the solutions attempted.238

2. Overcoming the Peer Monitoring Deficit

Second, the clusters should become a forum for performance monitoring. The cluster participants should be empowered to ask whether a camp-closure policy has benefitted the former residents, whether an entitlement to certain daily quantities of water is exacerbating conflicts over natural resources, or whether the benefits of housing multiple tribal communities in the same camp are being outweighed by the dissolution of existing communal ties. Specific questions such as these supplement the basic questions regarding the rights and daily needs of disaster victims. The challenge is that quantitative and qualitative performance standards are deeply contested among disaster responders and donors, with many agencies and observers arguing that performance monitoring leads to rigid response frameworks that provide aid according to indicators rather than according to real need.239 Only recently have aid agencies been able to agree on general standards for performance monitoring.240

An experimentalist framework, particularly one founded on independence, would require a much respect for autonomy and clearer policy statement on the desirability of departing from national and global performance indicators. Part III recalls how pressure from NGOs was unable to relax the rigid application of Sphere indicators in Chad. Strict adherence to quantitative indicators set at the global level is antithetical to an experimental approach. Under this model, clusters should become a forum for reevaluating the content and scope of existing performance indicators in the light of changing circumstances. This can work, as Janice Gross Stein points out, where networks are able to foster open discussion and "veil the face of power and the asymmetries of power" among their members.241

3. Reflexivity of Policy Making

Third, information about deviation and innovation should be able to flow up the chain, such that global standards may be modified in

238.

See U.K. FINANCIAL REPORTING COUNCL,

supra note 234, at 5-6

(discussing what constitutes a meaningful explanation).

239.

See generally Gross Stein, supra note 114 (discussing the challenge of

quantitative and qualitative performance standards).

240.

See Janice Gross Stein, The Politics and Power of Networks: The Accountability of Humanitarian Organizations, in NETWORKED POLITICS, supra note 122, at 151. 241.

Id. at 168.

light of local innovations.242 To date, according to most evaluations, the innovations that happen in local clusters tend to stay in local clusters.243 In some responses, where cluster recommendations were taken up at intercluster meetings, innovations did seem to have an effect across sectors.244 But information is often blocked between the subnational and the national levels, and the connection between the in-country clusters and the global clusters is close to nonexistent.245 It falls to outside consultants, performing reviews and evaluations, to uncover best practices and novel institutional forms. Experimentalist architecture requires a dramatically strengthened OCHA, which would be in charge of feeding information on shifting norms up and down the chain. In a perfect world of infinite time and resources, this might take the form of a database noting relevant standards, recognized challenges, attempted deviations, and results.246 In the real world of emergency response, some reflexive revision might be able lines of transmission, albeit with a worse signal-to-noise ratio. But as long as peer review stays locked in local clusters, the potential of the system to work fundamental changes in a way that could substitute for topdown accountability is practically precluded.

to take place with more

informal but efficient

The vertical movement of information should not only benefit the creation of better global standards but also further innovation along the horizontal dimension of the cluster system. A crucial aspect of experimentalist practice is that members of local units can take notice of emerging norms in other their adoption.247 In the reimagined cluster system, this may mean that

localities and urge

242.

The bidirectionality of reflexive rulemaking is emphasized by Dorf, supra note 210. This understanding embellishes on the concept of reflexive law developed by Teubner but disconnects this concept from the dogmatic tenets of systems theory. See generally Gunther Teubner, Substantive and Reflexive Elements in Modern Law, 17 LAw & Soc'Y REV. 239 (1983); Ulrich Beck, The Reinvention of Politics, in REFLEXIVE MODERNIZATION 1, 24-26 (Ulrich Beck, Anthony Giddens & Scott Lash eds., 1994) (discarding the strict logic of systems theory).

243. 244. 245. 246.

E.g., STEETS ET AL., supra note 10, at 52. See id. at 37 (discussing the DRC as an example). Id. at 32. Simon's description of the Toyota manufacturing floor exemplifies the promises and challenges associated with information pooling: "Elaborate displays visible from all points in the plant summarize what is happening at each station. When a problem that requires suspension of production occurs, its nature and location are communicated immediately to the entire plant. The premise is that, at the outset, we cannot say which people in the plant will have the knowledge and skills necessary to the solution." Simon, supra note 210, at 21-22. For humanitarians, the "display" would be the IASC and OCHA websites, and the "plant" might refer to a response in a single country. One should not, however, overlook the distributional effects of information technology, as many local NGOs do not have strong capacity in this area. See STEETS ET AL., supra note 10, at 62.

247.

See Dorf & Sabel, supra note 194, at 323 (describing the process of

"benchmarking" to invite comparisons across jurisdictions).

VANDERBILT]OURNAL OF TRANSNATIONAL

LAW

innovations in Kivu should be accessible to clusters operating in other regions of the DRC, but it also means that Haitian clusters should be able to consider programs developing in Pakistan. If information is pooled in a generally accessible way, it becomes possible for a member of a Chadian cluster to notice previously undiscovered pathologies in the response because these same problems have already been corrected in Sudan. Again, this aspect of the system is seriously limited by the present resource constraints on OCHA,248 but the expenses associated with information pooling may be significantly less than those needed to robustly manage the current hierarchical system of supervision.

4. Rethinking Hierarchy

The discussion

to this point has emphasized the horizontal aspects of experimentalist architecture. But top-down oversight-and even harsh sanctions-may be crucial to ensuring the track.249 In an experimentalist experimental process stays on humanitarian system, the vertical structure would not be eliminated, but it would be reconfigured to focus exclusively on ensuring the continued motion of the problem-solving process. The substantive aspects of the vertical relation would be jettisoned, with top-down interventions being limited to ensuring the appropriate level of disclosure, participation, and inclusion.250

that

By retreating from substantive issues, and even some procedural ones, the hierarchical structure of the cluster system might become less threatening the independence of non-UN humanitarian agencies. At the same time, more routine forms of mutual monitoring within the clusters could leave local actors with more tools to change policy at the ground level, and they need only seek redress at the next highest level of the hierarchy when the experimental process breaks down. In any case, whether or not the actual policy generated in these discussions matches global norms is far less important than ensuring that new policies are explained and subjected to continuous contestation from grassroots organizations and victims' advocates. This ideal is probably far more attainable in some sectors than others. But, importantly, experimentalism provides a coherent and intelligible theory for shaping the role of supervisory authority in a decentralized system.

248. 249.

See supra text accompanying notes 220-25. See Sabel & Simon, supra note 22, at 81 (discussing the incentive-based design of experimentalist regimes often including coercive norms such as "penalty defaults").

250.

See id. at 81-82 ("[An Experimentalist regime is ... more likely to impose [harsh monetary or criminal penalties] for failing to make or execute plans or to report performance.").

C. Participation as the Central Problem of a Reimagined System

tension,

addressing

autonomy-coordination

that is neither fundamentally democratic,

an experimentalist humanitarian system pins its legitimation hopes on the type, quality, and extent of direct participation by affected populations.251 This legitimation strategy demands a kind of participation its traditional representative sense, nor technocratic, but rather is deliberative and driven toward improving the quality and sensitivity of humanitarian action.252 Active rulemaking, peer monitoring, and reflexivity are not goods in themselves; they function as workable accountability mechanisms and legitimation strategies only to the extent that they correct for the dangers of institutional choice. They do this by ensuring a level of context sensitivity and responsiveness that would not necessarily be expected of humanitarian organizations on their own.253 While the relationship between victims and aid workers is not a new problem, participation re-emerges in this way as the central point of tension within on experimentalism.254

strategy based

a

This should in no way suggest that experimentalist structures solve the problem of participation. Indeed, ensuring the voice of affected populations in emergency relief, though stated as a central concern of most aid organizations, remains a perennial problem for

251.

Cf. Dana Brakman Reiser & Claire R. Kelly, Linking NGO Accountability and the Legitimacy of Global Governance, 36 BROOK. J. INT'L L. 1011, 1066-68 (2011) (stressing that the precise role of NGOs in an entity's legitimation strategy will determine the type of institutional mechanisms that will be necessary).

252.

To the extent experimentalism

is on associative or deliberative models of "democracy," rather than representative ones. See generally David Gartner, Beyond the Monopoly of States, 32 U. PA. J. INT'L L. 595, 62939 (2010) (discussing differences between these models).

is "democratic," the emphasis

253. Recall de Waal's critique that the "genuineness" of the humanitarian's commitment to effective relief delivery becomes pathological by generating an internationalized, technocratic culture that becomes desensitized to local realities. DE WAAL, supra note 181, at 4-5.

254.

To suggest that participation is key in this context is, not to fall into the trap of thinking that an NGO, or UN, activity is legitimized solely on the basis of its ability to "represent" the populations it serves. See Steve Charnovitz, Accountability of Non-Governmental Organizations in Global Governance, in NGO AccOUNTABILITY, supra note 113, at 21, 35-36 (debunking the problem of NGO representativeness as a "red herring"); Kenneth Anderson, 'Accountability" as "Legitimacy" Global Governance, Global Civil Society and the United Nations, 36 BROOK. J. INT'L L. 841, 873-77 against the representativeness of NGOs). Participation, rather, serves to counterbalance the relatively distanced perspective of the humanitarian international, in order to bolster or ensure the quality of the ideas being developed and adopted. The good idea is the central benefit of a working experimentalist structure, but local experience is central to the development of such ideas.

recent history of attacks

(noting

leveled

(2011)

/VOL. 47-239

humanitarian operations.255 Clusters have not necessarily made attempts to include and respond to affected populations any easier. Reviews have shown that cluster meetings are often held only in English and suffer from the jargon-heavy language of international humanitarianism, which is inaccessible not only to local populations but also to most national NGOs.256 The hope that clusters would be used to press for greater inclusion of local populations, in part because of their decentralized and collaborative nature, largely has not been realized.257 A recent study indicates that clusters have attempted to obtain information from local groups but that these same groups lack the power to influence cluster operations and generally fail to obtain funding for locally generated programs.258

If these problems cannot be rectified, then the experimentalist structure fails on its own terms. Experimentalism holds out the promise of involving all those affected by power to share in shaping and controlling it, both by opening initial participation and by transparently publicizing the results.259 In an analogous context, Dorf and Sabel note that the quality of service in public housing and community policing "depends so directly on the contribution of the beneficiaries that their active participation essentially makes them coproviders."260 The authors recommend a tiered structure, with intensive local participation at the lower levels and opportunities for civilian review at higher levels to address strategic issues.261 If, instead of involving the local communities in generating innovative solutions, clusters instead facilitate relatively sealed conversations among international relief actors, they risk increasing the divide between de Waal's humanitarian international and the populations they purport to serve.262

255.

STEETS ET AL., supra note 10, at 59 ("Thus, cluster work plans and strategies were in most cases not discussed with or validated by affected populations.").

256. 257.

Id. at 62; Miles, supra note 180, at 47. E.g., KAUFFMANN & KRUGER, supra note 42, at 39 (discussing the

accountability issues that exist with clusters).

258. 259.

Serventy, supra note 70, at 34. See Dorf & Sabel, supra note 194, at 288, 313 monitoring ... distributes authority from the 'rulers' to the 'people."').

("[L]earning by

260. 261.

Id. at 31-7. For an elaboration of experimentalism in community policing, see id. at 327-32. Public housing and policing are relevant analogies to humanitarian response; housing and safety are central concerns in the provision of aid, and, at a higher level of abstraction, each form of service provision reflects an enterprise where local conditions and concerns, such as crime, quality of housing, and humanitarian needs, are so bound up with local conditions and the knowledge of beneficiaries that their input becomes vital to developing best practices.

262.

See DE WAAL, supra note 181, at 65-85 (referring to the humanitarian international as "the international elite of the staff of international relief agencies, journalists, lobbyists and ...'conflict resolution' academics, consultants, specialist specialists and human rights workers"); cf. Anderson, supra note 154, at 890 (stressing the problem of an international civil society composed of global NGOs that serve "as

2014]1

Nor would ensuring participation necessarily solve the problem of participation. In an experimentalist structure, the question of who is a peer who deserves an invitation to participate must become a is constantly reassessed and perennial problem, which reevaluated in light of new information.263 Experimental structures must find ways to ensure that the very boundaries of the institution are open to contestation, lest the system recreate the insularity that to avoid. This becomes particularly important as it purported contemporary humanitarian responses may last for several years.264

itself

it may generate

In addition, effective participation may unlock new conflict zones that, under current arrangements, are effectively hidden. For conflicts between disaster-affected example, populations and their governments.265 A system that grants such voices direct access to the levers of humanitarian power might pose a significant threat to authorities, particularly in areas where the power of foreign humanitarian actors rivals that of the state itself.266 If the cluster approach does come to enjoy such success in securing participation, Will it come at the price of losing access to troubled areas of the world?

Experimentalism thus represents a fraught normative choice for the cluster approach. On the one hand, as accountability to affected populations remains a constant problem under any model of response, experimentalism holds out the promise of improving participation by devolving decisions to local groups and sensitizing actors to local context. On the other hand, 'without participation, the emphasis on peer review and information pooling risks further reifying the existing normative, cultural, and linguistic barriers between foreign and local actors.267 Reviews of the cluster system do not necessarily

their own gate-keepers," purporting to be accountable to an amorphous and curiously silent global populace).

263.

E.g., de Bdrca, Keohane & Sabel, supra note 187, at 786 ("Enlarging the circle of decision making, and keeping it accessible to new participants is a condition of success.").

264.

E.g., Sontag, supra note 2, at Al (reporting in advance of the three-year

anniversary of the Haiti earthquake).

265.

Cf. Dorf & Sabel, supra note 194, at 314 (arguing that experimentalism should "'(re)politicize' political institutions by introducing a novel form of deliberation based on the diversity of practical activity, not the dispassionate homogeneity of those insulated from everyday experience").

266.

E.g., Janil Lwijis, NGOs: What Government Are You?, in TECTONIC SHIFTS, supra note 180, at 69 (analyzing NGOs as international government at a local level); to HARVEY, supra note 164, at 3 ("The potential for international aid agencies undermine or inappropriately substitute for the state has often led to tense relations between states and international actors.").

267.

See JULIA STEETS & FRANCOIS GRONEWALD, IASC CLUSTER APPROACH EVALUATION, 2ND PHASE: COUNTRY STUDY-UGANDA 43 (2010) (stating that a topdown introduction of the cluster system "was detrimental to ownership compared to previous approaches and disempowered national and local actors").

indicate a framework has failed participation.

trend in either direction, suggesting only that the to alter the status quo of very little

VII. CONCLUSION

A recent wave of large-scale disasters has brought about a moment of transition in the life of international humanitarian institutions. The profile of the Haiti earthquake of 2010 has renewed calls for the accountability of aid institutions,268 and the cluster system itself has managed to sneak into mainstream news coverage, generally as the subject of criticism.269 The relatively high level of public attention provides an opportunity to reflect on the institutional arrangements through which disaster response is conducted and on the manner in which they exercise power. How one understands the problems associated with such power, and the means for its control, will guide the possibilities for institutional design and innovation.

The foregoing discussion should not be understood as making the strong claim that the rising salience of victims' rights or grassroots organizations has or will cause a shift away from state-centered modes of accountability. Rather, this Article finds that a horizontal accountability structure is emerging from the wreckage of a formal system that was continually unable to work, in practice or in theory.270 This emerging structure is not mere hardware; it is embedded with its own normative outlook for the system, which takes a different orientation than the formal structure. In this emerging orientation, victim participation is no longer merely one of many concerns facing the humanitarian enterprise. It is the keystone of the cluster system's legitimation strategy.

And it is a risky one. If, after Haiti, grassroots organizations are to publicly critique the international humanitarian emboldened enterprise and assess its responsiveness to local voices, then an accountability strategy that pins its hopes on context sensitivity and learning opens itself to strong and highly charged normative challenges.

268. Michael Jennings, International NGOs Must Address Their Accountability 2012),

Deficit, http://www.guardian.co.uk/global-development/poverty-matters/2012/feb/09/ngosaccountability-deficit-legal-framework.

POVERTY MATTERS

GUARDIAN:

BLOG

(Feb.

9,

269. 270.

See supra note 180 and sources cited therein. See also NGOs & HUMANITARIAN REFORM PROJECT, FIT FOR THE FUTURE? STRENGTHENING THE LEADERSHIP PILLAR OF HUMANITARIAN REFORM 15 (2010) ("in a system characterised by non-hierarchical relationships between partners, and strong competitive incentives for each of the partners, there should be little surprise that an accountability system based solely on hierarchical, vertical lines does not work").

More than simply failing

in its mission of inclusion, an experimental and self-correcting process can easily descend into a sealed-off, unresponsive form of peer interaction.271 Indeed, the pathologies of network-style interactions among an international elite class of experts seem particularly dangerous in the context of humanitarianism. Recently, Brian Concannon and Beatrice Lindstrom have emphasized that the aid effort in Haiti was partially undermined because the United Nations and other bodies "extensively and inappropriately relied on international NGOs to be the voice of the people."272 This type of error is troublesome in any case, but it is absolutely fatal to an institution whose strategy for accountability and legitimacy rests on cognitive openness and sensitivity. To the extent that the cluster system shifts toward an experimentalist framework, it will be haunted by images of victims' groups being barred from cloistered compounds or sidelined by inaccessible jargon.273

The renewed emphasis on victim participation expressed here, which is associated with the development of horizontal structures for policy experimentation and peer review, is one view of the future of humanitarian institutions. By focusing on nascent developments in practice, this Article has attempted to draw out their implications for the transformation of international disaster response. But these developments are by no means foretold. In light of this investigation, experimentalism, in the guise of horizontal accountability, remains a promising of humanitarian institutions, but it is not without its own normative tensions. Having interrogated these pitfalls, this Article has not solved the problem of humanitarian organizations, but it has presented a clearer map of the paths ahead.

reorganization

approach

271.

For similar critiques from various perspectives, see generally Philip Alston, The Myopia of the Handmaidens: International Lawyers and Globalization, 8 EUR. J. INT'L L. 435 (1997) (emphasizing the backroom character of networks); Sol Pieciotto, Networks in International Economic Integration: Fragmented States and the Dilemmas of Neo-Liberalism, 17 Nw. J. INT'L L. & BUS. 1014 (1997) (making an argument similar to Alston's); Kenneth Anderson, The Ottawa Convention Banning Landmines, the Role of International Non-Governmental Organizations and the Idea of International Civil Society, 11 EUR. J. INT'L L. 91 (2000) (emphasizing the nonrepresentative, elite character of international NGOs).

272.

Brian Concannon Jr. & Beatrice Lindstrom, Cheaper, Better, LongerLasting: A Rights-Based Approach to Disaster Response in Haiti, 25 EMORY INT'L L. REV. 1145, 1177 (2011) (internal quotation omitted).

273.

See Miles, supra note 180, at 49 ("An opportunity to set a good example for the rest of the world instead established a model of how best to act against the interests of a devastated people and render them invisible.").

$@

--- FOOTNOTES ---

1.

30.

See HUMANITARIAN RESPONSE REVIEW, supra note 19, at 46-52 (setting

'Service Clusters'

_________

'Cross-cuttingissues'Co

Cyclone Nargis

Medical Corps

Sources: STEETS ET AL., supra note 10; ROBERT TURNER ET AL., INTER-AGENCY REAL

TIME EVALUATION OF THE RESPONSE TO CYCLONE NARGIS (2008); INTER-AGENCY

STANDING COMM., GUIDANCE NOTE ON USING THE CLUSTER APPROACH TO STRENGTHEN

See generally Generic Terms of Reference for Sector Leads at the Country [IASCI, HEALTH CLUSTER GUIDE: A INTER-AGENCY STANDING COMM. Level, PRACTICAL GUIDE FOR COUNTRY-LEVEL IMPLEMENTATION OF THE HEALTH CLUSTER, at Annex A, 1-3 [hereinafter IASC, Cluster Lead TOR] (describing the role and duties of lead agencies).

64.

On this concept, see ANNE-MARIE SLAUGHTER, A NEW WORLD ORDER 247-

65.

66.

143.

results, if it can be relied upon at all. INTER-AGENCY STANDING COMM., SURVEY ON THE CIVIL-MILITARY GUIDELINES AND REFERENCES FOR COMPLEX EMERGENCIES (July 1, 2010).

175.

See RANDOLPH KENT, HUMANITARIAN FUTURES PROGRAMME, MAPPING THE MODELS: THE ROLES AND RATIONALE FOR THE HUMANITARIAN COORDINATOR 6-10 (2009) (noting the strengths of the multi-hatting approach).

181.

ALEX DE WAAL, FAMINE CRIMES: POLITICS AND THE DISASTER RELIEF

INDUSTRY IN AFRICA 3-4 (1997).

196.

See also GLYN TAYLOR ET AL., ALNAP, STATE OF THE HUMANITARIAN

198.

DAVID RIEFF, A BED FOR THE NIGHT: HUMANITARIANISM IN CRISIS 19 (2002)