UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Measuring human security : a new view of Cambodian vulnerability Owen, Taylor Reid 2003

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_2003-0631.pdf [ 6.78MB ]
JSON: 831-1.0099710.json
JSON-LD: 831-1.0099710-ld.json
RDF/XML (Pretty): 831-1.0099710-rdf.xml
RDF/JSON: 831-1.0099710-rdf.json
Turtle: 831-1.0099710-turtle.txt
N-Triples: 831-1.0099710-rdf-ntriples.txt
Original Record: 831-1.0099710-source.json
Full Text

Full Text

MEASURING HUMAN SECURITY: A NEW VIEW OF CAMBODIAN VULNERABILITY by TAYLOR REID OWEN B.A. Bishops University, 2000 A THESIS SUBMITED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES (Department of Geography) We accept this thesis ^ asxQnforrning to the required standards THE UNIVERSITY OF BRITISH COLUMBIA SEPTEMBER 2003 © Taylor Reid Owen 2003 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of (h< JQ<?P^ ' T A K ^ The University of British Columbia Vancouver, Canada Date DE-6 (2788) Abstract Human security is a new and contested concept. Although gaining legitimacy in many academic and policy communities, it has no single accepted definition, no universal foreign policy mandate and no consensus-commanding analytic framework for its measurement. This is in part do to a perceived conceptual ambiguity coupled with an inherent paradox in its measurement - the broader the spectrum of human security measured, the more difficult data collection and aggregation become. This paradox has forced the six existing measuring methodologies to be either broad and conceptually accurate, but of questionable feasibility and reliability, or narrow and feasible, but not representative of the full range of insecurities. In response to this difficulty, a measuring methodology is proposed centered around a new perception of space. Selecting indicators based on their regional relevance and aggregating them using their common denominator, location, allows the methodology to be conceptually broad, analytically accurate and practically feasible. Using a Geographic Information System, this methodology documents threats, allows for analysis of spatial correlations, and provides an invaluable tool for policy makers in regions of high vulnerability. This methodology was tested through a case study in Cambodia. 13 human security threats were established and spatially referenced local data were collected detailing them. An overlay analysis of high threat regions in each of the 13 threats revealed 'hotspots' of insecurity and a correlation analysis revealed a significant relationship between high landmine contamination and high poverty, dengue fever, HIV/AIDS, domestic violence and gun injuries. In addition, poverty was spatially correlated with dengue fever, domestic violence and landmines. While these correlations do not imply necessary causality, they do show a degree of significance that warrants further inquiry n Table of Contents Abstract i i Table of Contents i i i List of Tables iv List of Figures and Map vi Glossary of Acronyms vii CHAPTER 1 Introduction: Human Security and Geography 1 CHAPTER 2 Conceptual Background 6 2.1 Philosophic Root 7 2.2 From Traditional to Human Security 9 2.3 Human Security Defined: From Narrow to Broad 10 2.4 Human Security Questioned: Critiques and Rebuttals 15 2.5 Measuring Human Security: Progress Report and Problematic 21 2.5.1 Why Measure Human Security? 21 2.5.2 Who is Measuring and How: A Review of Measuring Methodologies 33 2.5.2 The Measurement Paradox 34 CHAPTER 3 A New Methodology For Measuring Human Security 36 3.1 Towards a New Definition of Human Security 37 3.2 From Definition to Measurement : 39 3.2.1 Stage One: Threat Assessment ...40 3.2.2 Stage Two: Data Collection and Organization 41 3.2.3 Stage Three: Data Analysis and Visualization 43 3.3 Summary 47 CHAPTER 4 Cambodia Case Study 50 4.1 Why Cambodia? ; 50 4.2 Cambodia Background 51 4.3 Spatial Orientation 54 4.4 Implementation of Methodology 58 4.4.1 Stage One: Threat Assessment 58 4.4.2 Stage Two: Data Collection and Organization 60 4.4.3 Stage Three: Data Analysis and Visualization 69 Data Visualization 69 Data Analysis 76 4.4.2 Conclusions of Analysis 93 CHAPTER 5 Ways Forward 95 5.1 Human Security and Geography 95 , 5.2 Implications of the Cambodian Case Study for Human Security 95 5.2.1 Theoretical Implications 95 5.2.2 implications for Measuring Human Security 97 On the Appeal of Human Security 97 On the Procedure of Measuring Human Security 99 Works Cited 102 Appendix 1 Comparing Measuring Methodologies Chart 107 Appendix 2 Letter and Chart for Cambodian Expert Researchers 110 Appendix 3 Stage One and Two Process Chart 112 Appendix 4 Root Causes of Conflict 113 Appendix 5 Data Tables for 13 Security Threats ; 114 Appendix 6 Base Maps and High Threat Maps for 13 Human Security Threats 121 List of Tables .1.1 Traditional vs. Human Security 7 1.2 Measuring Problems and Proposed Solutions 35 3.1 Example of Stage 2 Table 43 3.2 Example of Stage 2 Table 44 3.3 Example of Stage 2 Table 2 46 4.1 Example of Interview Database 59 4.2 Metadata Table 63 4.3 Example of Data Table 64 4.4 Sample of 115,000 Record US Bombing Data Table 66 4.5 Example of 'Hotspot' Table 77 4.6 Correlation Table 1 80 4.7 Correlation Table 2 83 4.8 Correlation Table 3 84 v List of Figures and Maps 2.1 HDIvs.IHI 27 3.1 Stage One Diagram 41 3.2 Diagram of Causal Relations 47 4.1 Map of South East Asia 54 4.2 Map of Cambodia '• 55 4.3 Maps of Province and Regional District Boundaries 56 4.4 Maps of Commune Boundaries and Village Points 57 4.5 Cambodia Threats in Security Categories 60 4.6 Map of 115, US Bombing Points 67 4.7 Diagram of Spatial Joining Process 70 4.8 Diagram of the Creation of High Threat Map 73 4.9 Diagram of Data Unions Creating Final Data Table 75 4.10 Hotspot Map 78 4.11 Radarsat Flood Map 85 4.12 Outline of Peak Flood Level 85 4.13 Overlay of Peak Flood Level and Poverty Severity 86 4.14 Example of Road Map 88 4.15 Example of Road Map Zoom 88 4.16 Example of River Map 89 4.17 Example of River Map Zoom 89 4.18 Example of Topographic Map Overlay : 90 4.19 Example of Topographic Map Overlay Zoom 1 90 4.20 Example of Topographic Map Overlay Zoom 2 91 4.21 Example of Topographic Map Overlay Zoom 3 91 4.22 Landmine Fields 92 4.23 Landmine Fields Zoom 93 vi A c r o n y m s CDRI Cambodian Development Research Institute COMFREL Committee for Free and Fair Elections in Cambodia CPP Cambodian Peoples Party FUNCIPEC United Front for an Independent; Neutral and Co- operative Cambodia GDP Gross Domestic Product GECHS Global Environmental Change and Human Security GIS Geographic Information System HDI Human Development Index HPI Human Poverty Index ICBL International Convention for the Ban of Landmines ICHISS International Commission for Humanitarian Intervention and State Sovereignty ICRC International Commission of the Red Cross IHI Individual Human Insecurity IHS Index of Human Security IO International Organization LIGI Liu Institute for Global Issues NGO Non-Governmental Organization PLG Partnership for Local Governance PYHS Population Years of Human Security UN United Nations UNDP United Nations Development Program UNTAC United Nations Transitional Authority for Cambodia UXO Unexploded Ordinances WFP World Food Program WHO World Health Organization YIHS Years of Individual Human Security vii Chapter 1 Introduction: Human Security and Geography Human security is a new and contested concept. Although gaining legitimacy in many academic and policy communities, it has no single accepted definition, no universal foreign policy mandate and no consensus-i commanding analytic framework for its measurement. For many this is of little concern, that 'human security' was the coalescing force behind the International Convention to Ban Landmines and the International Criminal Court is enough to prove that it is both representative of popular sentiment and legitimate as a tool of international policy making. For others, however, this lack of clarity, poses a fundamental problem. If human security is anything that could harm anyone, then its policy mandate and analytic utility are rendered useless - by being overly inclusive human security risks becoming meaningless. The following thesis addresses the central criticism of human security, its ambiguity. This is done by looking at human security with a distinctly geographic perspective. By introducing the notion of space, it is possible to see human security not as a ubiquitous catch all term, but as an analytic concept with a specific meaning in a specific place. Human security in one location means something very different than human security in another. Only by acknowledging this, however, can we truly understand what human security as a whole entails. This thesis argues that the process of spatial clarification is imperative for the concept to remain prevalent both as a critique of the dominant security paradigm, traditional, or, realist security, and as a viable framework for foreign policy direction, development planning, and the study of causal and correlative relationships between social, environmental and economic threats. Central to the debate over the utility of human security is the feasibility of its measurement. This thesis centers on the measurement of human security, but assumes that broadly articulating the parameters of human security - particularly at the sub-national level - must predicate this process. Consequently, a brief history of the concept is provided, a definition is proposed, and its critics concerns are addressed. 1 Second, the thesis argues that measuring human security - despite its critics' concerns - is a worthy academic exercise. It introduces the six existing methodologies for measuring, concluding that they are all hindered by an unaddressed paradox: the more accurate conceptualization of human security used, the less representative the measurement becomes. Inversely, the more a measurement tries to represent all of the threats effecting individuals, the less feasible it becomes. In Chapter 3, a methodology is proposed that explicitly addresses this paradox of accuracy versus feasibility. Using an inclusive conceptualization, intra-state data, and the framework of a Geographic Information System, it is possible for a measurement to be at once conceptually broad, practically feasible and analytically relevant. In Chapter 4, the methodology is rigorously tested through a Cambodian case study. Questions of data quality and availability are addressed and each stage of the process is actualized. The result is an interactive human security mapping tool capable of locating hotpots of human insecurity and revealing spatial correlation between threats. This tool will be returned to policy makers, development practitioners and academic researchers in Cambodia. The Geographical Precedent Although the methodology proposed in this thesis is entirely original, both the underlying concept of human security and the practice of spatially aggregating varying types of data is common to the field of geography. As an academic discipline geography is necessarily broad. Including, for example, economic, physical, social, gender and urban geography, the field is clearly well positioned to address a concept as interdisciplinary as human security. More specifically, the specializations of hazard geography and risk geography closely parallel the underlying assumption of measuring human security - that it doesn't matter what harms are afflicting a region, but simply that people are at risk. 2 The parallel fields of hazard identification and risk assessment offer a constructive precedent for measuring human security. Rooted in human ecology, hazard research originated as means of analyzing environmental extremes, namely, floods (Barrows, 1923). As with human security, researchers were looking at both a conceptual and practical problem; conceptually, the societal relationship to flood location (Kates and Burton, 1986) and practically, how this translated into better flood management policy (White 1945, 1964). In this early research, hazards were firmly rooted in the analysis of physical events. It became clear, however, that a hazard could also be a socially constructed situation (Cutter et al, 2000, 714), an event not rooted in the physical environment, but as a consequence to human actions, such as technology, or, technological failure, (Kates et al, 1985; Sorenson et al., 1987; Cutter, 1993; Mitchell, 1998). This shift, led to a contextualization of hazard research. A hazard was now seen as very much rooted in a social, political, temporal, organizational and, spatial context (Cutter et al., 2000: 715). Perhaps more importantly for this thesis, is that this led to the development of methodologies incorporating both empirical and social analysis (Mitchell et al. 1989; Palm, 1990). The study of this interplay, between hazard and societal context, is the focus the study of risk and of vulnerability. Risk assessments are grounded in the mathematical study of probability and confidence intervals (Covello and Mumpower, 1994: 33). Although based on the precedent of mathematical inquiry, the first risk assessments per se, attempted to link an accident, originally seen as random, with a probability of occurrence, moving the interplay from pure chance, to having a degree of risk1. Where as the study of hazards focuses on the event, and the study of risk focuses on the probability of interplay between the event and its impact, vulnerability addresses the nature of the impact itself. The study of vulnerability is based on a combination of exposure and lack of resilience (Cutter et al. 2000: 716). What is particularly important for this thesis is that this interplay is location specific. What one is exposed to as well as to some extent, ones ability to mitigate the threat, is spatially determined (Droughts-Wilhite Kasperson's social amplification of risk model pioneered the modern study of risk (Kaperson et al., 1989). 3 and Easterning, 1987; Land Degradation - Blakie and Brookfield, 1987; Severe Environmental Risk -Kasperson et al., 1995). This has led to what Cutter and Solecki (Cutter and Solecki, 1989) call a hazard-of-place. The interplay of social, political and economic factors- interacting separately, in combination with one another, and with the physical environment- creates a mosaic of risks and hazards that affect people and the places they inhabit (Cutter et al. 2000: 716) Curiously, while recognized, this mosaic has only once, and quite recently, been quantitatively assessed. Cutter et al. (Cutter et al, 2000) provide a practical methodology, the hazard-of-place model of vulnerability, for combining varying biophysical and societal vulnerabilities and have conducted a case study for a county in South Carolina2. Their methodology, however, being somewhat narrow in both its theoretical base and spatial focus (only conducted for one county), is only a very early trial of what is sure to be a growing field. In addition to conceptual parallels, the methods of Geography are also relevant to the study of human security. Broadly, looking at space, or, location as a common attribute, provides a new variable with which human security can be assessed and analyzed. As will be suggested in Chapter 3, adding this common attribute provides a method of overcoming many of the difficulties of measuring such a broad, and some would argue ambiguous, concept. Practically, Geographic Information Systems (GIS), provide the means to spatially analysis human security data. Mapping and spatially analyzing data provides a powerful analytic tool and visualizing what would usually be seen as complex tables and graphs is a very effective tool of advocacy (Denis, 1991, 2001; Richardson, 1999; Spence, 2001) It should be noted that this case study was found at the final stages of writing this thesis. The full extent of its parallels have not been revealed and 4 As a relatively new concept in critical security studies, human security has commanded noteworthy academic attention and unprecedented public policy resonance. This being said, clear conceptualization, accurate measurement and useful analytical inquiry are essential to its survival in an international climate increasingly hostile to non-traditional security. This paper seeks to add to the growing debate over what security entails and to provide a tool to study and implement broad based human security policy. 5 Chapter 2 Conceptual Background: History, Concept and Measurement Human security, although a new and contested concept, has a deep historical root, a strong theoretical base and has generated an ongoing discussion on its measurement. In opposition, however, the pervasive nature of the competing realist paradigm, a degree of definitional uncertainty, and the inherent paradox of its measurement, threaten to relocate the concept to the fate of previous broadening (critical) security concepts. This chapter will briefly outline the historical context of human security, review the spectrum of definitions (from broad to narrow) and rebut its dominant critiques - providing a conceptual introduction and rationale for its measurement. The six proposed measuring methodologies will then be reviewed with particular attention to the paradox of conceptual representation vs. analytical accuracy. It will be concluded that this paradox can be reconciled by reorienting the role of space in the analysis. Acquiring data at the sub-national and aggregating variables using their common attribute, space, directly addresses the paradox that limits the other six measuring methodologies. Human security is naturally related to more traditional concepts of security derived from military analysis of the international relations paradigm. While the dominant traditional security paradigm relies on an anarchistic relationship among states, human security shifts the reverent object to the individual. A security threat, therefore, entails not simply something that can threaten the integrity of the state, but of the individual. This clearly opens the discourse to a much wider array of possible threats, but in so doing brings to bear difficult questions of conceptual clarity and practical feasibility. The concept was formally introduced to the security discourse as a post Cold War, 1989, response to threats deemed outside of the capabilities of traditional security. Power balances were effective in dealing with the nuclear threat, but would they be capable of providing security from intra-state conflict, environmental disasters, famine, disease, landmines, and domestic violence? 6 \ As seen in Table 1.1, the major difference between human security and more traditional approaches to security is, therefore, this broadening of focus. Table 2.1 Traditional vs. Human Security Type of Security Referent Object Responsibility to Protect Possible Threats Traditional Security The State The Integrity of the State Interstate War, Nuclear Proliferation, Revolution Human Security The Individual The Integrity of the Individual Disease, Poverty, Natural Disaster, Violence, Landmines, Human Rights Abuses 2.1 Philosophic Roots Emma Rothschild provides a succinct summary of the intellectual history of human security, analyzing that while this is a contemporary concept, its roots can be discovered in a century-long debate, from the Treaty of Westphalia (1648) to the Concert of Europe (1815). At its core, argues Emma Rothschild, Human Security is a comparably undisciplined argument for a return to enlightenment liberalism (Rothschild, 1995: 54). It is only one of many reactionary questionings of the traditional security paradigm, one that has yet to fully define itself, let alone compete with a centuries old debate. Indeed, many of the core principles of human security are crude reflections of Montesquieu, Rousseau, and Condoorcet, but so too are principles of state security rooted in intellectuals of the time - Kant, Hobbes and De Grotius - whose opposing state centered world view arguably prevailed over these more pluralistic beliefs. Since the debate over the relationship between individual and state as the reverent object of security is not new, some comment on past perspectives is useful3. As Rothschild points out, one side of the security debate of the 18"1 Century is rooted in pluralist beliefs focusing on the protection of the individual. For Montesquieu, this was a singular focus on freedom and 3.The dichotomous nature of this debate is worth noting. A pattern, seen in the traditional security vs. human security debate as well as with the broad vs. narrow conception of human security itself. As I will argue, only by returning to the core protection of the individual, from all serious threats, will we be fulfilling our societal responsibilities of protection. 7 the perceived rights of individuals over the dictated security provided by the state (Ibid: 61). Security for Adam Smith meant the protection of the individual from "sudden or violent attack on one's person or property" (Ibid: 62). This security, he continues, is the most important prerequisite for a successful and "opulent" society. Similarly, Condorcet described a societal contract in which the security of the individual was the central principle. Fear of fear, explains Shklar, was the principal enemy to a liberal society (Shklar: 21). If freedom from fear were not guaranteed, then individuals could not be effective members of a political relationship. This liberal perspective was widespread, but not unanimous. Although in agreement over the vital role of individual safety, others believed that this could best be achieved as a consequence of the security of the state. The state, thus, acts as protector from both external and internal threats. For Hobbes, it meant little whether man's insecurity, 'solitary, poor, nasty, brutish and short' was at the hands of a local thief or an invading army. Protection from either, he believed, was the absolute responsibility of the state. For this protection, the citizen should give up any and all individual rights to his country, his protector (Ullman, 1983: 130). As arguably the birth of realism, Hobbes put forward a model of absolute state control over the security of individuals- security prevailing over liberty (Hobbes, 1651). While also looking at the role of the state in providing individual security, Kant envisioned a higher authority still. He proposed a universalist international order, based primarily on the moral guidance of it member nations, on a common good, a global society. (Haftendorn, 1991: 6) As a middle ground between the two, De Grotius proposed a more moderate international dynamic, one not guided by supranational law, but by a balance of power amongst states. The mutual interest of independent but coexisting state entities, therefore ensures the security of all (Ibid: 7). Although each gave rise to a different school of international thought (Hobbes and realism, Kant and global security, De Grotius and international security (Ibid: 7)), all based the primary responsibility of protecting individual security at the hands of the state. This would become the dominant world view, arguably winning out over more liberal thinking, until the end of the Cold War. 8 2.2 From Traditional to Human Security Do you know, I said, that governments vary as the dispositions of men vary, and that there must be as many of the one as there are of the other? For we cannot suppose that States are made of "oak and rock," and not out of the human natures which are in them, and which in a figure turn the scale and draw other things after them? Plato, The Republic Human security's ostensible roots can be found in early liberal philosophic writings. Its practical manifestation, however, is representative of a post Cold War skepticism toward the dominant tradition security paradigm. Whether Hobbesian, Kantian or Grotian in origin, the dominance of the traditional, state centered security paradigm reached a peak during the cold war. For 40 years, the major world powers entrusted the security of their populace, and to a certain extent of the world, on a balance of power among states. For this prevailing Realist view, the referent object of security is the state and presumes, in a very Hobbesian fashion, that if the state is secure, then so too will those that live within it (Bajpai, 2000: 32). This type of security relied primarily on an anarchistic balance of power (power as the sole controlling mechanism), the military buildup of two superpowers and on the absolute sovereignty of the nation state4. States were deemed in the scholarly literature and security analysis to be entirely rational entities, with the maximization of power guiding national interests and policy (Kim, 2001). Security was seen as protection from invading armies, protection was provided by technical and military capabilities and wars were never to be fought on home soil, rather, proxy-wars, were used if direct combat were necessary. With the fall of the Berlin Wall, and the subsequent string of successful secessionist movements, the realist view of international relations and security came into question. A l l of a sudden, the traditional controlling mechanism of the Westphalian state system no longer seemed entirely reliable. The world was now more fractured, and the new players could not always be trusted to play by Cold War rules. Often, as in the extreme cases of the Serbian government's attack on the Kosovar Albanians and the Rwandan Hutu's 4 The most prominent theoretical approaches that work with this definition of security are neo-realism (e.g. Waltz, 1979; Mearsheimer, 1994/95), neo-liberal institutionalism (e.g. Keohane, 1984; Keohane and Martin, 1995) and constructivism (Wendt, 1992; Newman, 2001) 9 genocide of the Tutsis, the state was responsible for the insecurity of the very people it was meant to protect. What has become clear is that despite the macro level stability created by the east-west military balance of the Cold War, citizens were not safe. They may not have suffered from outright nuclear attack, but they were being killed by the remnants of proxy wars, the environment, poverty, disease, hunger, violence and human rights abuses. Ironically, the faith placed in the realist world view, and the security it provided, > masked issues threatening the individual. Once the central foci of security (Enlightenment liberalism, as discussed earlier in the chapter), the protection of the person was all too often negated by an over-attention on the state. Allowing key issues to fall through the cracks, "traditional security" simply failed at its primary objective: protecting the individual. This new type of instability led to the challenging of the notion of traditional security by such concepts as cooperative, comprehensive, societal, collective, international and human security (Baylis, 1997). Although these concepts move away from a focus on inter state relations, human security takes the most dramatic step by making the referent object not the state, society or community, but the individual. 2.3 Human Security Defined: From Broad to Narrow The human security concept, therefore, shifts the referent object of security from the state to the individual. This is necessary due to the aforesaid failures of the traditional security paradigm, and also in order to direct research and policy towards actual issues threatening peoples' lives. As a general description, Rothschild (Rothschild, 1995: 55) describes human security as one of four changing places of security. Beginning with the state, she sees security being brought down to the individual, up to the international system or supranational physical environment, across (broadening) from a focus on the military to include the environment, society and economy, and finally, the responsibility to ensure security diffused in all directions to include local governments, international agreements, NGO's, 10 public opinion, forces of nature and the financial market. Although not an explicit definition, this conceptualization provides an example of both how narrow the traditional paradigm has been, as well as how complex the expansion of the concept can become. Although many attempts have been made to more specifically define what is an inherently ambiguous concept (as it by definition encompasses a potentially unlimited list of threat), two conceptual schools of thought have emerged in which most definitions can be grouped. These are the broad and narrow, or, "freedom from want" and "freedom from fear", conceptions of human security5. Human Security: Broadly Defined6 "Security" says Kofi Annan, "can no longer be narrowly defined as the absence of armed conflict, be it between or within states. Gross abuses of human rights, the large-scale displacement of civilian populations, international terrorism, the AIDS pandemic, drug and arms trafficking and environmental disasters present a direct threat to human security, forcing us to adopt a much more coordinated approach to a range of issues." (Annan, 2000) Most of the definitions of human security are rooted in the broad, or United Nations Development Program (UNDP), school of thought. Although critics rightfully point to a potential ambiguity from grouping so many threats under one heading, conceptual clarity emerges if three key attributes are considered; its scope 5 It should be noted that I personally feel that this categorization is far too simplistic and not very accurate. Not just because the most of the literature is based in the broad conception, but because many of the definitions used in the literature on human security incorporate elements of both want and fear. Also, while some definitions might be broad in that they stress human development priorities, they may in fact still be very narrow in the scope of the threats they include - such is the case with the King and Murray conception (King, 2000). A large amount of literature has emerged on Human Security (see Krause, 2000), a full review falls out of the reach of this project. For the three best overall literature reviews on human security, see Alkire, 2001 and Hampson, 2001, and the Harvard Program on Human Security at http://www.cbrss.harvard.edu/programs/hsecurity/hspapers.htm . I have chosen what I believe are the seminal scholarly works in the field and those that best represent broad, narrow and critical perspectives on human security. 6 A spectrum has been used to describe the possible definitions of human security. It can be seen in its broad sense as widely incorporating a long list of possible threats, from traditional security threats such as war to more development oriented threats such as health, poverty, and the environment. In its narrow sense, the spectrum, although still focused on the individual, and therefore incorporating many more threats that traditional security, it is limited to violent threat. This cold include landmines, small arms, violence, and intra-state conflict. 11 of coverage, its system based approach to understanding causal relationships7, and its focus on the vital core of the individual. These three critical aspects of broadly defined human security are exemplified by the UNDP (UNDP, 1995), Salbina Alkire (Human Security Commission, 2002) and Jorge Nef (Nef, 1999) respectively. First, the U N D P conceptualization establishes human security's broad scope. The 1994 UNDP Human Development Report is generally seen as the first significant attempt at articulating the broad approach to human security. This report describes human security as having two principal aspects; the freedom from chronic threats such as hunger, disease and repression, coupled with the protection from sudden calamities. The report concedes that it is broad, but explains that that this is simply a reflection of the number of significant harms that go unmitigated. As a conceptual structure, seven components of human security the U N D P offers: economic, food, health, environmental, personal, community and political security8 (UNDP, 1995). 1. Economic security: the threat is poverty 2. Food security: the threat is hunger and famine 3. Health security: the threat is injury and disease 4. Environmental security: the threat is pollution, environmental degradation and resource depletion 5. Personal security: the threat includes various forms of violence 6. Community security: the threat is to the integrity of cultures 7. Political security: the threat is political repression9 What is important about this categorization is that it sets the "boundaries of the tent" very broadly, clearly separating itself from past security re-conceptualizations. Also, it forces other definitions of human security (outlined below) to justify their narrowing from this very broad starting point. Second, Jorge Nef points out the importance of a components based approach to defining human security (Nef, 1999: 13-26). He describes five interconnected subsystems of human security: ecosystem, economy, society, polity and culture. For Nef, these five are all in complex interplay, their linkages defining the I will outline one definition that falls into each of these categories, for other definitions of human security see Hampson, 2002. 8 These 'components' will become the theoretical foundation of my measurement methodology, considered in Chapter 3. 9 As summarized in Mack, 2002. 12 nature of systemic balance. More crudely, this points out that if causality is going to be addressed, then the system boundaries must be set very broadly in order to capture all of the possible significant variables. Acknowledgment of the interconnectedness of human security components, however defined, is absolutely critical both to understanding causality and properly addressing policy. A third significant attribute of the broad conception of human security is its focus on the vital core of the individual. This is essential in order to separate 'human security' from 'human development', a term which is more linked with wellbeing than dire emergencies10. The focus on the individual in human security can perhaps be best illustrated by Alkire (Alkire, 2002), in the following quote. 'The objective of human security is to safeguard the vital core of all human lives in ways that enhance human freedoms and human fulfillment".11 Key to this definition is the focus on all critical and pervasive threats to the vital core. Instead of providing a laundry list of threats, she sets criteria, surpassing which, any issue becomes a threat to human security. She also presumes that although institutions cannot be expected to protect people from all harms, they should at least address those that unnecessarily take lives. With these three attributes in mind, the broad conception of human security begins to clarify. According to proponents of this definitions of the human security term, it must be inclusive, casting a "broad tent", it must separate its components into different types of security in order to address causality, and it must set a threshold demarcating the vital core in order to separate itself from human development. Human Security: Narrowly Defined On the other end of the spectrum addressing definitions of human security is the 'narrow', or what has become known as the Canadian Approach. By using a definition that primarily focuses on violent threats, 1 0 This concept has been articulated by Salbina Alkire, most recently in the Human Security Commission Report 2002. 8 Initially put forth by Alkire in "Conceptual Framework for Human Security 16 February, 2002" 13 the Canadian Approach clearly separates human security from the much broader and already established field of international development. Indeed, the Canadian government acknowledges the UNDP conception as merely a phase in the development of human security, but envisions a much more focused definition, one centered on violent threats, as an instrument of policy (Acharya, 2001). i The Canadian definition, therefore, restricts the parameters of human security to a focus on the threat of violence to the individual. This can come from a vast array of threats, including the drug trade, landmines, ethnic discord, state failure, trafficking in small arms etc. This must, as former foreign minister Lloyd Axworthy points out (Axworthy, 2001), be countered primarily by the use of soft power, such as diplomatic resources, economic persuasion, and the use of intelligence and information technology. The Human Security Report (first issue to be published late 2003), to be produced by the Human Security Center at the Liu Institute for the Study of Global Issues, University of British Columbia, and is modeled after the Human Development Report, also uses a narrowly defined understanding of human security, limiting its scope for pragmatic and methodological reasons (Mack, 2002: 4). For instance, pragmatically, the Human Development Report already covers the freedom from want side of the spectrum, so another such report would be redundant. Methodologically, understanding the relationship between underdevelopment and violence necessarily requires a separation of the dependant and independent variable (Mack, 2002)12. While a vigorous debate over definitions and ambiguity in the human security paradigm is perused in the literature, a strong argument for the narrow conception is simply the number of successful international initiatives using its parameters. In fact, most of the significant policy advances achieved in the name of human security have used this narrow definition. For instance, the International Conventional for the Ban of Landmines, the International Criminal Court, as well as the recent international NGO focus on child soldiers, small arms and armed groups, have all been undertaken using a Canadian (narrow) perspective. 1 2 I will argue that this ambiguity is only true of the broad conception when all components are aggregated together. If kept on their own, all under the heading of human security threats, then meaningful correlation is possible. In fact, the very fact that they are all deemed human security threats, forces a degree of comparison that might otherwise go unnoticed. 14 Despite principally aligned world views, proponents of the broad and the narrow definitions of human security have yet to come up with a single consensus commanding definition. This is problematic, as it has led to four fundamental critiques (below). In order for human security to have a meaningful impact, at least within the academic community, its proponents must agree upon a single definition and end what is a self destructive debate. 2.4 Human Security Questioned: Critiques and Rebuttals Quite apart from a heated debate over the definition of human security, there is a remarkably strong questioning of, and resistance to, the very existence of the concept, particularly by scholars from the realist view of international relations. Instead of squabbling among themselves, proponents need to counter these substantial denigrations, as divisions over definitions are only fueling criticisms. While a significant aspect of this response should include a consensus on what human security is, a direct rebuttal to specific worries is also imperative - sustained ambiguity will only lead to further critiques. The prevalence of critiques included within the human security literature cannot be underestimated. Even proponents of the term highlight what could potentially be fatal flaws in the reasoning of the concept. Although some simply parlay an 'old boys club' 1 3 rhetoric for the status quo (realism), and others exemplify a weak grasp of what human security entails, some put the very theoretical and practical foundations of the concept into question. Principal criticisms of human security as a valid departure from traditional approaches to thinking about security can be roughly grouped into four concerns14: 1 3 Critics of human security, often from the realist school of international relations, very often argue what is a very ideologically conservative view of the world. 1 4Another criticism, somewhat out of the context of this analysis, is Human Security as a western construct (Macfarlane: 1). A brief comment on the influence of the realist security paradigm on present 'southern' nations' security views should be enough to quell this line of thought. The very reason that some nations 15 Critique One- The Definition is Unclear This is by far the most common criticism, one that addresses the operational and policy connections of the paradigm of human security (and, I would argue, is the easiest to both make and defend), and takes several forms. Macfarlane, Paris and Bajpai, Foong Khong and Deudney all iterate the following concern: Shifting the referent object of security to the individual, implies that any threat to its (the person) ability to exist constitutes a security threat. Subsequently, this results in an unmanageable laundry list of threats, exemplifying a lack of conceptual clarity and practical feasibility. Bringing everything into the security rubric dilutes the concept to practical stagnation. Put simply, says Macfarlane (Macfarlane, 2000: 1), human security has no boundaries, putting its analytic utility into question. Paris (Paris, 2001: 88) goes a step further and comments that human security does not even have a definition, and, that much like sustainable development, "everyone is for it, but few have an idea of what it means". Conceptualizations, he continues, tend to be expansive and vague, offering little guidance to the policy community. Following this reasoning, the policy implications of making everything a security issue are of concern. Foong Khong (Foong Khong, 2001: 1) points out that issues are securitized in order to put them at a higher policy priority, "it becomes an urgent issue, worthy of special attention, resources, and fast track or immediate amelioration or resolution, perhaps by military means". If all the possible concerns encompassed in human security are securitized, he argues, our ability to prioritize is rendered useless, and nothing will receive the attention it deserves. This will simply lead to false hopes (Ibid: 2-3). Macfarlane (Macfarlane, 2000: 1) uses similar reasoning, and suggests that issues that would be prioritized under the human security rubric, might better be handled by "more appropriate points of view and by more appropriate institutions". This argument was in fact also used by Deudney in his seminal critique of 'environmental security' (Deudney, 1990). are hesitant to endorse human security, is their fear of being taken advantage of by the powers of traditional security, not their moral positioning in poverty, disease, violence, environmental disasters etc. 16 What is prevalent throughout this line of reasoning is a fear that policy priorities will run out of control if the human security mandate proliferates. 'How can we (the security community of the western world) possibly help everyone', proclaim scholars advocating traditional security, 'when we are already busy looking after ourselves (national security)?' There are three clear answers to this "narcissistic" concern: First, the list of human security threats, especially in the developing world, is so vast, precisely because 'we' do not properly address them. Second, 'we' have the ability and capacity to deal with a vast array of security threats. A slight re-designauon of military budgets, for example, would more than cover most of them15, and our institutions are filled with bright people developing innovative solutions to the practical difficulties of these complex situations16. Third, it's in our own national self interest to do so. It is becoming increasingly clear that our security is inextricably linked to the security of other nations and consequently, other individuals. Critique Two- The Role of the State is Unclear? A second line of criticism of the human security paradigm involves a type of treatise to the state. These types of criticisms - also inherent in the work of MacFarlane, Krause and Foong Khoon - seem to extend beyond simply a defense of realism, or traditional security, and focuses on impassioned epitaphs on all of the good that the state has done for world and individual security. MacFarlane (MacFarlane, 2000: 1) comments that the critique of the state inherent in the individualistic approach of human security is "overdone". Proponents of human security, he argues, should simply look to the state, and the promotion of democratic principles, as a means of addressing their concerns. Krause (Krausse, 2000: 24) takes a similar line in pointing out the required re-conceptualization of sovereignty should human security principles take root. He worries that adjusting procedures of humanitarian 1 5 The UNDP has stated that $US 20Billion would solve most of Africa's food, water and health emergencies. This is a fraction of the US$ 390 United stated annual defense budget. 1 6 Although a sufficient debate of this last point is out of the context of this thesis, I would argue that the huge intellectual capacity in universities and development oriented research institutions is being wasted by a dualistic process of researchers not writing "policy friendly" literature (Mack, 2002) and the policy community not engaging researchers for consultation on the concerns that they simply don't have the time to reason out. intervention, as prescribed by the more decentralized approach of human security, will not necessarily solve all world problems. Foong Khoon, citing Axworthy, (Foong Khong, 2001: 3) challenges the very root of human security, claiming that it is based on the false assumption that the security of the individual will necessarily ensure international security - the primary goal of realist security. This however misreads Axworthy's principal point, that the safety of the individual is "integral to the promotion and maintenance of international peace and security" (Axworthy, 2001: 19), not an unconditional cause of. Never are such absolute propositions suggested. I contend that all of these criticisms, and many others like them, demonstrate a poor understanding of the concept of human security. Never have the type of absolute positioning that seems to get accredited to proponents of the concept, been suggested. Nowhere in the literature, either academic or policy, has the abolishment of the state ever been prescribed. Neither has the state ever been discounted as a means of protecting people from certain harms. Proponents of human security argue that the state alone, based on a 350 year history, has proven itself incapable of protecting against all harms. Subsequently, we should be looking for both conceptual and practical tools to help us address those issues that the state either is not able, or not willing, to deal with. Alternative ideas to the traditional security paradigm, such as those recommended in the International Commission for Humanitarian Intervention and Interstate Sovereignty (ICHIIS, 2001) and the International Convention to Ban Landmines, ideas that address the changing dynamic of humanitarian intervention and global relations, do not significantly compromise state sovereignty and are in the spirit of human security. Critique Three- In Whose Interest is Unclear? A n interesting historical criticism of human security addresses its conception into the security discourse and its association with the rise of 'middle power' 1 7 influence. 1 7 A term often used to describe countries without large militaries but that exert international influence in other ways, such as through international organizations, development assistance and diplomacy. Canada is seen as an exemplary middle power. 18 Paris rightfully points out that human security is potentially expansive and vague. However, he wrongly associates this with the expansive and vague values and foreign policy interests of the Human Security Network1 8 member states. However, in rebuttal it can be argued that the list of human security threats is expansive and vague because of the many complex problems that fall under its mandate, not because someone or some nation has chosen them. Krause (Krause, 2000) makes the point more directly, attributing self interest as cause for both the original U N D P conception of human security "to influence policies of states in the preparatory phase of the 1995 U N Conference on Social Development" (Krausse: 8), as well as for the members of the Human Security Network, who "discovered a lucrative set of foreign policy issues and an opportunity to champion a 'new diplomacy' that enhanced their role on the world stage" (Krausse, 2000: 20). He goes further, citing an "identity crisis" of the three most important member states, Canada, Norway and Switzerland. Krause's argument itself has many contradictions, however, as he then backtracks and separates the creation of the concept (self interested) from the content of the concept. The content he asserts, is not self interested, as Canadians are not dying from landmines and child soldiers (Ibid: 21). This type of criticism appears to be rooted in a realist premise that state actions must be self interested. It also ignores the possibility that addressing human security concerns abroad could be in the interest of the state and be both mutually self interested and altruistic. Further, it should be noted that while realist security is by definition self interested, its proponents often use its more humanitarian consequences as legitimizing factors. For human security to be criticized on the same grounds, therefore, could be said to be hypocritical. Critique Four- The Nature of Causal Analysis is Unclear? Meaningful causal analysis requires the clear separation of dependant and independent variables. Critics of broad conceptions of human security claim that including many variables under one heading blurs these 1 8 The Human Security Network is an organization of like minded state, who together are advocating a human security approach to foreign policy. See www.humansecurity.org 19 lines and makes analytic inquiry next to impossible. "If everything that causes a decline in well-being is labeled a security threat, the term loses any analytic usefulness and becomes a loose synonym for 'bad'" (Deudney: 464). 'The study of causal relationships", says Paris, "requires a degree of analytical separation that the notion of human security lacks" (Paris, 2001: 93). He continues that by combining physical threats with socio economic threats under the rubric of human security, analytic inquiry is rendered impossible as both variables are both dependant and independent of the final condition. Similarly as a proponent of the narrow conception of human security speaking against the broad, Mack (Mack, 2002) states that: If the term 'insecurity' embraces almost all forms of harm — from affronts to dignity to genocide -its descriptive power becomes diffused. A concept that explains everything in reality explains nothing. Moreover, in order to examine relationships between, for example, poverty and violence, each must be treated separately for the purposes of analysis. Any definition that conflates dependent and independent variables renders causal analysis virtually impossible (Mack, 2002). Both Paris's and Mack's comments, appear not to counter any of the arguments proposed in the human security literature. Proponents of human security tend not to argue so dogmatically that all possible threats need to ultimately be labeled the same (human security) or given the exact same policy priority, but rather that as serious threats to large numbers of people, they all share a common attribute. In fact, putting them all under the same heading helps to facilitate the very type of causal analysis that both critics rightfully desire. Under the broad umbrella of human security, are many possible threats, some interconnected some independent, it is for these very social scientists to decipher these important correlations. Although a rebuttal to all four of these critiques is reasonably straightforward, their prevalence and veracity accentuate the need for clarity within the discipline. One way in which this can be established is by clearly defining and then measuring human insecurity. An inherently difficult process, measuring human security has been attempted, with varying degrees of success, using 6 different methodologies. 20 2.5 Measuring Human Security: Progress Report and Problematic Increasingly central in the debate over the utility of human security is the feasibility of its measurement. This section analyses the six existing methodologies for measuring human security, showing that the narrow conception of human security outlined in Human Security Report, with its limited number of indicators and narrow definition of the concept, is perhaps the most feasible. This, however, reveals a paradox: that the more conceptually accurate a measurement is (i.e. includes all harms affecting individuals) the less analytically feasible it becomes. 2.5.1 Who is Measuring and How: A Review of Measuring Methodologies Six broad frameworks have been proposed in the literature for measuring human security.19 They can be distinguished from one another in three ways: how they define the concept of human security, what they propose to measure, and the methodology they use to aggregate and analyze the data. 2 0 1. Generalized Poverty (a) Definition Used Gary King and Christopher Murray 2 1 (King, 2000) define human insecurity as a state of 'generalized poverty'. This state exists when a human being ranks below a predetermined threshold in any domain of well-being. They argue that there are important qualitative differences in well-being or life experience, above and below a pre-defined level. If any individual or group falls below its threshold, a minimum quality of life, it is said to be in a state of generalized poverty. Implicit in this definition is that no one indicator, if below the threshold, is any more significant than another. For instance, this approach solves the 1 9 These six were chosen because they make a clear reference to measuring 'human security'. There are many measuring methodologies that analyze similar variables and could be considered measurements of human security type threats. It would be useful for a more extensive study to look at these other methodologies. 2 0 It should also be noted that some of the measures are simply hypothetical, (Human Security Report, Human Security Audit, Generalized Poverty), where as others have been tested or operationalized (GECHS, Leaning and Arie, Rummel). 2 1 Gary King is a Professor of Political Science at the Kennedy School of Government, Harvard University, and Christopher Murray is a consultant with the World Health Organization. 21 problem of weighting and leaves the establishment of thresholds as the only subjective aspect of the process. (b) What is Measured King and Murray define the domains of well-being as income, health, education, political freedom and democracy, closely paralleling the United Nations Development Program (UNDP) definition2 2. For each domain, a single dichotomous indicator is chosen on which individuals can score 1 or 0. Interestingly, they do not include a measure of violence, although the domains chosen were all deemed to be 'things worth fighting for' (King, 2000: 14-15). Indicators Used (c) Methodology Three methods have been proposed by King and Murray to further apply their theory of generalized poverty. The first measure is the Years of Individual Human Security (YIHS). This represents the expected years that an individual will spend outside of a 'generalized state of poverty'. The latter exists if an individual scores zero on any one of the above indicators. The use of language becomes a little unusual in this measurement scheme. For example, scoring zero on 'political freedom' would put an individual in a state of 'generalizable poverty' even though that individual might be quite affluent. 'Generalizable poverty', in other words, does not necessarily equate with poverty in the traditionally understood sense of the term. The second measure is 'Individual Human Security' (IHS). This represents the proportion of an individual's lifespan that she or he could expect to spend outside of a state of generalized poverty, insuring 2 2 The U N D P definition of human security was introduced in the 1994 Human development report (UNDP, 1994). Including 7 categories (economic, food, health, community, environmental, political and personal) it is considered the broadest possible conceptualization of human security. It is often criticized for too closely resembling what are usually considered "development' rather than "security" concerns (King, 2000). Domain 1. Income Indicator GNP per capita in international $'s converted to purchasing power parity Quality of health scale Literacy rate or average years of schooling Freedom house, measure of societal freedom. The fraction of adults able to participate in elections 2. Health 3. Education 4. Political Freedom 5. Democracy 22 a more accurate control for age. Finally, a method of aggregating the YIHS for a particular population -within a nation-state, for example, has been proposed, 'The Population Years of Human Security' (PYHS). This may be useful in the study of development and security policies (King, 2000: 10-13). 2. The Human Security Audit (a) Definition Used For Kanti Bajpai 2 3 (Bajpai, 2000) human security is defined as the protection from direct and indirect threats to the personal safety and well-being of the individual. This is derived from commonalties between the Canadian and U N D P conceptions of human security and combination of safety and freedom is highlighted (Bajpai, 2000: 52). (a) What is to be Measured Indicators Used Direct Threats 1. Local 2. Regional 3. National 4. International Violent Crime, abuse of women/children Terrorism, genocide, government repression Societal violence, international war, banditry, ethnic violence Interstate war, weapons of mass destruction, land mines Indirect Threats 1. Societal Level- Lack of basic needs, such as food, water and primary health care Disease Employment Levels Population growth or decline Natural Disasters 2. Global Level- Population Movement Environmental Degradation Unequal Consumption (b) Methodology The first stage of Bajpai's methodology is to measure the potential threat to the individual. This is done by collecting quantitative data for an array of potential direct and indirect threats. (See above list) Although Bajpai claims that the data for this stage is abundant, there may be some problems with continuity and 2 3 Kanti Bajpai teaches at the School of International Studies, Jawaharlal Nehru University, New Delhi and is presently a Visiting Fellow at the Joan B. Kroc Institute of International Peace Studies, University of Notre Dame. 1 2 3 accuracy given the broad range of indicators. A problem in this approach is that much of the data he requires is either aggregated from sparse and questionable data sources or simply isn't available for the developing world. Moreover, judgments about potential threats - as against actual physical harm - are necessarily conjectural and are unlikely to command consensus. Second, Bajpai seeks to measure the capacity of the individual to cope with potential threats. This is done through a qualitative assessment of peoples' or governments' capacities. Possible indicators might include government anti-racism policies, as opposed to incidents of racist abuse (Bajpai, 2000: 53-56). Although this notion of threat vs. capacity methodology is interesting, it has some practical limitations. For instance, the method by which indicators would be aggregated is not explained and the weighting between the threats and the capacities appears to be entirely subjective. In addition, there is no assurance that the capacities will be directly relevant to the threats posed. If they are not, then any process of aggregation will have little meaning. For example, a country may have a robust social welfare system (a strong indicator of capacity to deal with an economic threat), and so consequently it would score a high capacity ranking. This same country, however, may have no effective disaster response program and so be highly vulnerable to the threat of major natural disasters. 3. The Global Environmental Change and Human Security Project (GECHS) (a) Definition Used The G H E C H definition of human security builds on the premise that certain environmental and social conditions, when coupled with increasingly vulnerable societies, may lead to insecurity of the individual. Security in this context is only achieved when individuals have the option, physically and politically, to end or adapt to threats to their environmental, social or human rights. This methodology attempts to measure a broad range of human security threats with a focus on environmental components. 24 The GECHS definition of human security puts significant focus on a cumulative causality between the physical environment and personal safety. They also point out potential indirect threats stemming from either an environmental condition, or the human response to that condition (Aviso, 2000:2). (b) What is Measured Social, environmental, economic and institutional domains of security are the focus of the GECHS Index. Four indicators for each domain (see below) are then selected using the following criteria. • Relevance to the selection framework • Existence of a theoretical or empirical relationship • General availability of the data • Data commensurability • Adequacy of the spatial coverage There is, however, no mention of a prioritization of these factors. As with Bajpai's methodology, if their relative priority is based on an entirely subjective reasoning, the soundness of the methodology should be questioned (Aviso, 2000: 3-5). Indicators Used Social Urban population growth Young male population Maternal mortality ratio Life expectancy Environmental Net energy imports Soil degradation-Safe water Arable land Economic Real GDP per capita GNP per capita growth Adult literacy rate Value of imports and exports of goods and services Institutional Public expenditures on defence versus education Gross domestic fixed investment Degree of democratization Human freedom index 25 (c) Methodology Once indicators are selected, the methodology involves three stages. First, data are collected for all indicators at the national level with a detailed a time series (1970- 1995) as available. Missing data are estimated using linear regression. Second, data are standardized into a common scale2 4. Finally, the index is calculated using cluster analysis, assigning each a degree of severity (insecurity) between one and ten to each indicator for each country (Lonergan, 2000). The Index of Human Insecurity (IHI) is the only one of the four proposed indicies that has been actualized using real data. Interestingly, there is a strong linear relationship between the IHI and the Human Development Index (HDI). This is perhaps not surprising since both are seeking, in effect, to measure human development. The IHI simply uses many more indicators - the 16 being somewhat analogous to the core three indicators of the HDI. Fig. 2.1 HDI vs. IHI a 1.5 S ex <D o -a u *, 0.5 h o 1 3 5 7 IndexofHuman Insecurity (Aviso: 2000) But the GECHS approach raises several questions. First, what is the difference between development and security as defined by GECHS? This question is at the centre of the definitional debate about human security. If the list of harms under human security becomes too exhaustive, it quickly begins to resemble those found under the concept of development. This has the potential to render both its analytic and policy 2 4 The details of this step are not clear from the literature, yet they appear crucial to the validity of the final measurement. 26 utility redundant. In this regard, it is also possible that the results of the IHJ, being so similar to the HDI, could be used to support a narrower definition of human security such as the one used in the Human Security Report - if broad measurements are redundant, narrow definitions will prevail. Second, what differentiates the IHI from the HDI? Is it telling us anything new and if not, why use it? It could also be asked if either index tells us anything more than the single indicator of infant mortality, which is often thought to be a good proxy for the HDI . 2 5 4 . The Human Security Report (a) Definition Used The Human Security Report, at U B C ' s Human Security Centre uses the most restrictive measure of human insecurity, limiting it for pragmatic and methodological reasons to deaths caused by armed conflict and criminal violence. This is a much narrower spectrum of threats than those embraced by any of the other measuring methodologies (see Appendix 1). The Report does not propose mapping injuries from war and criminal violence due to lack of reliable data for most countries, but does suggest that number of deaths per annum deaths might be a good proxy for this (HSR Proposal, 2000). (b) What is Measured The Human Security Centre has not yet decided whether or not to include a composite Human Insecurity Index in the Report. If so, the most likely indicators would be armed conflict deaths and homicides. The convention already used for deaths from disease, natural disasters and criminal violence - i.e. number of deaths per 100,000 per year in a particular country would be used. The data from each would be aggregated on a national and possibly a regional level. The following three defenses of the GECHS index are presented by the creators. The first is that the HDI index is intended to achieve a stable indicator and therefore might not catch the degree of detail present in the 16 indicators of the IHI. Second, that the IHI has a much stronger theoretical link to the concept of human security and development. And third, the IHI attempts to deal with some issues of perception by incorporating some qualitative data. Again, these three statements would have to be further defended, and hopefully will be in a forthcoming GECHS publication (Aviso, 2000). 27 Like all others, this index also confronts a number of difficulties. Currently no data are collected on the absolute numbers of deaths per year, as against the number of armed conflicts that exceed a certain death threshold26. Second, the data are subject to a variety of biases. Although there is no reason to assume that they are all skewed, the data are almost certainly not accurate enough to have any confidence in detailed individual country rankings - a difference of one or two places on the individual country ranking having little substantive meaning. This problem is similar to that faced with the Human Development Index of the Human Development Report. Ranking countries in quintiles may, however, be possible - generalized groupings providing proxy for overall security. Third, while criminal violence (unlike the armed conflict) data are collected by governments, it too is subject to inaccuracies, often being both under-recorded and under-reported. National collection of data are often subject to political bias. Moreover, data are often published several years after it has been collected - making it difficult to use in an annual index. (HSR Proposal, 2000: 2-5). While the Gary/Murray and GECHS indices do not include violence the Human Security Report Index would not include deaths caused by disease, starvation or natural disasters. The Human Security Report argues that there are important relationships between underdevelopment and human insecurity (narrowly defined), but that to explore them each must be treated separately for the purpose of analysis. Indicators Used 1. Battle-Related Deaths in Armed Conflicts per 100,000 of population collected by Uppsala University. 2. Homicides- Collected by national governments per 100,000 of population. (c) Methodology Since there is a common scale for both indicators - number of deaths - aggregation of the two measures is unproblematic. When total death toll data from armed conflicts are collected (they are not collected at the moment) it will be possible to measure battle-related deaths per 100,000 of population, allowing for more meaningful national comparisons. The Human Security Centre is currently negotiating with the university of Uppsala to collect these data. 28 Although this methodology is certainly feasible, the approach's strong link with more traditional, militaristic notions of security, and its very limited scope could be criticized for straying too far from the root of human security. 5. Harvard School of Public Health (a) Definition of Human Security Jennifer Leaning and Sam Arie at the Harvard School of Public Health have proposed a measuring methodology that is unique in both its weighting and its conception of human security (Leaning and Arie, 2000). They define human security as the necessary social, psychosocial, economic and political aspects of human life that allow the individual to survive in times of acute crisis or chronic deprivation. For this measurement, human security signifies resilience. (b) What is Measured Leaning and Arie (Ibid, 2000), propose a composite, capability based methodology focusing on two domains of human security: Minimum Material Inputs and Basic Psychosocial Supports. Material Inputs encompass the necessities for minimum levels of survival (water, food and shelter) protection from life threats. They suggest that this could be captured by using the Human Poverty Index (HPI), an inverse measure of the Human Development Index, using the same dimensions of education, health and economic life (Ibid: 11). Basic Psychosocial Needs entails the basic needs from integrity, recognition, participation and autonomy. This is captured by looking at three interdependent variables: Relationship with location (sustainable sense of home and safety), relationship with community (a network of constructive social or family supports) and relationships with time (an acceptance with the past and a positive grasp of the future) (Ibid: 17) Indicators Used 1. Minimum Material Inputs 29 Human Poverty Index 2. Basic Psychosocial Supports Relationship with Location Relationship with Community Relationship with Time (c) Methodology Citing the difficulty in measuring vulnerability, Leaning and Arie suggest measuring the negative indicators of insecurity. They point out that the HPI already does this, covering one half of their measurement. For the concept psychosocial supports they offer three measures: dislocation (for home); dynamic inequality between groups (for community); and high discount rates (for a positive sense of the future) (Ibid: 30) There is no indication given to weighting, aggregation or empirical methods, although for the indicators themselves, a mix of qualitative and quantitative data is suggested. Much as with the Human Security Audit, the feasibility and final accuracy of this measure is questionable. As many would question the appropriateness of even including psychosocial threats under human security, most would contest their equal weighting to all other possible threats. Also, the presumption that HPI captures this other half of the spectrum (including violence, corruption, environmental disasters, disease etc.) is difficult to comprehend as it is only intended to represent poverty. The three dimensions of the psychosocial threat are interesting, however, and could be included in a "broad" measurement methodology, perhaps as one of several domains. 6. Rummel: Testing Whether Freedom Predicts Human Security and Violence (a) Definition Rummel states that "Human security is a general concept including the human and economic development of a people, their level of wealth and prosperity, and the threat to their lives by genocide, mass murder, war and political turmoil and instability" (Rummel, 2001: 3). His measurement, by contrast, is looking for a narrower and simpler deduction. In fact, based on his methodological structure, human security should first 30 be defined by its components, violence, human development, and economic development and could after his final analysis, be simply equated with "freedom". (b) What is Measured The measure divides human security into three domains: violence, economic development and human development. To measure each of these concepts, a long list of indicators is used. In order to carry out the final comparisons of freedom to human security, another list of freedom indicators is used, classified under civil rights and political liberties. Indicators Used Human Security Violence Economic Development Human Development Freedom Civi l Rights Political Liberties (c) Methodology Using a 190 country data set for 1997-1998 data, Rummel has used a factor analysis on a list of possible indicators for each of the three domains of human security27. The purpose of this analysis is to determine the accurate relevance and weighting of each possible measures based on its statistical prevalence in each nation. It is possible that one measure, or a complex combination of measures, can be said to be an accurate representation of the domain. Overall, comparable scores for each domain are then calculated and aggregated. The same is then done for the freedom measure. For a description of 'factor analysis' see "Understanding Factor Analysis" by Rudolf Rummel http ://w ww.mega, nu: 8080/ampp/rummel/ufa. htm 31 The second part of the analysis seeks to determine whether or not freedom, as the fundamental premise of a liberal democracy, either predicts or causes human security. This is examined by Rummel using the following three analyses: 1) Include freedom with the other score for human security and carry out a common factor analysis. 2) A contingency analysis of different levels of freedom versus levels of insecurity. 3) A regression analysis of the human security scores onto those measuring freedom (Rummel: 16). The conclusions from these three steps are: 1) The measure of freedom, or enslavement, of a people is the common factor in their human security or insecurity. 2) The human security of a nation is contingent on the freedom of its people. 3) Freedom is basic to human security- the more freedom people have the greater their human security (Rummel, 2000: 18-21). Aside from telling us something about the relationship between freedom and human security, this type of analysis can significantly enhance development strategy. For example, it could be concluded that funding for human rights organizations in a country of high insecurity could be an effective means of combating much wider societal insecurities. 2.5.2 The Measurement Paradox This brief review of the six types of human security analysis reveals how contested - not to say confused -contemporary treatments of human security are. The six methodologies are based on quite different conceptions of human security. For instance, whereas GECHS stresses environmental variables and King/Murray stress developmental issues, neither addresses the measurement of violence. The Human Security Report index, on the other hand, includes neither environmental nor development factors in its measure of human security. The chart in Appendix 1 displays the indicators used by each methodology and demonstrates how much of the broad UNDP definition each one incorporates. 32 What becomes clear if this chart is paralleled with the feasibility of each methodology, is that if attempts are made to broaden an index by including more indicators, issues of weighting, timeliness and accuracy become increasingly problematic28. This results in a difficult paradox: The more conceptually accurate -broad- a methodology attempt to be (i.e. closer to representing all possible threats), the less practically and analytically feasible it becomes. As a methodology expands its conceptualization of human security, closer to the broad UNDP definition, it become increasingly difficult to both aggregate and differentiate between each method's autonomous variables. In addition, particularly on a global scale, the data simply are unlikely to be available to fill out a "laundry list" of threats for every country. This leads to either significant gaps when comparing one country to another, or the use of old, problematic and unreliable data. One solution for proponents of broader concepts of human security man that of the Human Security Centre at U B C may be to simply use numbers of deaths from a range of different causes as indicators of human insecurity. Such an approach would avoid problems of aggregating and weighting, since 'death' can be said to be a common denominator. Death tolls are also usually a good proxy indicator for lesser forms of physical harm. To make country comparisons, national mortality data from various causes would be recorded in terms of deaths per 100,000 per year - this is already done in the health realm and with criminal violence data. As has been discussed, The Human Security Report uses such a methodology, but only seeks to measure the physical violence or 'freedom from fear' side of the spectrum. It may, however, also be possible to use deaths to measure a broader conception of human insecurity, embracing more varied threats to human well For extensive discussion on the problems of data collection see, Collier, 2001; Brauer, 2001; Mial, 2001; Mack: 2002. 2 9 A proposal for this Morality Index can be found in (Owen, 2002) 33 Another solution to the measuring paradox is to measure human security at the sub national level, using space, its geographical unit of analysis, as the common, comparable denominator. A methodology for such an analysis is conceptualized, proposed and tested in the following chapter. 2.6 Summary The traditional notion of security, rooted in the protection of the state and relying on an anarchistic balance of power, has proven to be insufficient in addressing the majority of problems people face around the world. This discrepancy was highlighted and exasperated by the end of the cold war. In order to address those events, conditions and actions actually harming people, one proposed alternative, human security, shifts the referent object of security from the state to the individual. Where as a security threat was once only something that threatened the integrity of the state, under the human security rubric, it is anything that threatens the integrity of the individual. While this arguable more accurately addresses the majority of harms, it also poses some difficult analytic and policy problems, viz. how does one distinguish and prioritize threats if all harms are security concerns. In response to this problematic, six proponent of human security have attempted to measure insecurity. In their attempts, however, they invariably run into problems of data availability, integrity and aggregation, resulting in the following paradox; that the broader a methodology attempts to be (the closer to the broad notion of human security), the more difficult accurate measurement becomes. This has resulted in methodologies either being narrow and feasible or broad and impossible to implement and/or inaccurate. The following three solutions will be provided in the following chapter to address the three roots of the measurement paradox. o 34 Table 2.2 Measuring Problems and Proposed Solutions Problems Solutions Data Availability Only measure regionally relevant threats Look sub-nationally for data Data Integrity Accept subjectivity, mitigated by local knowledge and disciplinary experts Data Aggregation Use space as a common denominator Geographic Information Systems (GIS) Absolutely integral to this re-conceptualization of human security measurement is the notion of space. No index has yet to stray for national level indicators. By shifting our perception of space, and measuring insecurity at the local level, I will argue that a much more meaningful representation may be achieved and the problems of measuring paradox overcome. 35 Chapter 3 A New Methodology for Measuring Human Security Human Security has a deep historical tradition, an academic resonance and a practical prescience. Despite this, proponents cannot agree on a single unifying definition, or on a single empirical methodology for its measurement. Faced with formidable opposition human security could well go the way of other critical security concepts30: This chapter will detail a methodology that addresses the principal concerns of both the definitional and measurement debates; viz. the paradox of accuracy vs. feasibility and bring a distinctively geographical interpretation of human security into the debate. While broad measurements are capable of capturing all harms threatening the individual, narrow ones more concisely articulate the nature of some of these threats. After reviewing previous attempts to operationalize the human security concept, it can be argued that what is needed is a definition and measuring methodology that is bom representative of the broad conception, yet empirically accurate as the narrow conception. The primary focus of the following methodology is to accurately and justly represent the concept of human security, remaining practically broad for practitioners and conceptually accurate for academics. This is done by beginning first with a broad, UNDP notion of human security, subdividing the concept into six categories; economic, health, food, political, personal and environmental. Country specific threats in each category are then determined and local, spatially referenced data collected. Finally, each threat is mapped, highest threat regions are isolated and all maps are overlaid to determine hotspots of aggregated human insecurity. comprehensive, international or co-operative security. 36 3.1 Towards a New Definition of Human Security Although many governments and organizations have adopted the concept of human security to broadly direct foreign policy directions, few have explicitly defined it31. But as explored in chapter 2, of the definitions that have emerged, none has garnered significant widespread acceptance32. For the methodology used in this thesis, I will use a hybrid definition. With the goal of remaining both broad and concise, the definition recognizes that there is no difference between a death from a flood or from a gun, all are considered threats to human security, but it must also be able to separate and categorize all of these possible threats, human security gets sub-divided into six threat categories. The definition, therefore, takes two parts. The first is derived from The Human Security Commission (HS Commission, 2002): Human security is the protection of the vital core of all human lives from critical and pervasive threats. The advantage of this definition is that it remains true to the broad nature of human security, while clearly separating it from more general concepts of human wellbeing and development. Making the referent object 'all human lives' both focuses on the individual while also indicating a universalism in its mandate. As the highest human insecurity is likely to occur in the developing world, this is particularly important. The threshold for what is deemed a human security threat is set by the terms 'vital core' and 'critical and pervasive threats'. This is important in order to ingrain a necessary degree of severity within the concept. 3 1 As suggested earlier, this is due to a hesitancy to articulate what is thought to be a fuzzy concept, a reluctance to position themselves explicitly on the spectrum of conceptualizations, as well as the difficulty of defining a concept with such broad parameters. 3 2 For a concise list of definitions see the Harvard School of Public Health http://www.hsph.harvard.edu/hpcr/events/hsworkshop/comparison_definitions.pdf 37 The vital core, as The Human Security Commission points out, is what constitutes a minimum level of survival (HSC: 4). Survival is a basic attributes that make us human. Threats to the vital core threaten lives. Reference to 'critical and pervasive threats' establishes both severity and immediacy. As there are an unlimited number of possible threats, only the most serious, those that take or seriously threaten lives, are included. The second part of the definition used in his thesis addresses the issue of conceptual clarity. It establishes clear categories under which all human security threats are ordered. These categories are not threats themselves, but rather are conceptual groupings, providing a degree of disciplinary alignment to what is an overarching concept. Individuals require protection from environmental, economic, food, health, personal and political threats. By grouping all possible threats into six. categories, I argue that human security becomes both more manageable and analytically useful. These categories are based on the original UNDP definition discussed in Ch. 2 (UNDP, 1995)33. This is important as the UNDP's definition is the most widely sited definition of human security and is considered the origin of the concept. The final definition is therefore; Human security is the protection of the vital core of all human lives from critical and pervasive threats. Individuals require protection from environmental, economic, food, health, personal and political threats. It should be noted that the seventh UNDP human security category 'community security' was omitted from my definition. This was done because I feel it conflicts with the first part of the definition, limiting human security to critical and pervasive threats to the vital core. I don't feel that integrity of culture, while undeniably important, fits under this conception. 38 I argue that perhaps most importantly, this definition is dynamic. It refrains from simply listing threats, recognizing that no possible list can be conclusive, and that it is the protection of the individual that should be the focus. 3.2 From Definition to Measurement Although the above definition articulates a concept and categorisation of human security, by itself it does little to reveal exactly what the threats are, who they are effecting and where they are located. Without this qualifying information, human security means nothing. With this in mind, I have developed a methodology, based on the structure of a Geographic Information System (GIS) that facilitates the collection, organization and spatial analysis of human security information. This methodology, incorporating vast amounts on interdisciplinary data, takes a subnational approach and is capable of clearly identifying the human security threats of a country, isolating 'hotspots' of aggregated insecurity and determining important spatial relationships amongst rarely compared security threats. This methodology also directly takes on the paradox of measurement: that the broader the conceptualization, the more difficult and less accurate the measurement. It does so by addressing one of the key difficulties of complex system analysis, the aggregation of differing data types. By building the methodology around the spatial reference of data, a common denominator is created - space - allowing for direct aggregation and analysis without creating a subjective nominal scale. Also critical is the methodologies focus on identifying local hotspots rather that creating a national index. Although many global indexes point policy makers to underdeveloped or insecure countries, none, isolate specific regions and problems within these countries. This arguably contributes to poor, nationally ubiquitous development policy and a lack of meaningful correlation analysis between harms. 39 3.2.1 Stage One: Threat Assessment As noted in Ch. 2, by shifting the focus of vulnerability from the state to the individual, human security attempts to incorporate what are traditionally considered development concerns into the realm of security studies - focusing on threats such as natural disasters, communicable diseases, dire poverty, human trafficking or landmines. This shift in focus, however, presents a potentially unmanageable mandate. Critics rightly point out that if all potential harms were included, the concept would be rendered analytically and practically useless. This has led many to define human security by simply listing what they feel are legitimate threats to the individual - the laundry list. I feel that this is the wrong approach as it can be argued that insecurity should not be defined by the presence of a hypothetical list of threats, but rather by what threats affect peoples' lives based on empirical research in real situations. Following this approach, human security threats will (and necessarily should) be unique to every country - different areas being afflicted by different harms. Therefore, no one list of threats can properly define human security34. The first stage of the methodology, therefore, is to determine from grounded empirical research, what specific threats affect a particular country or region. This can be achieved in a number of ways. Ideally, regional experts in each of the six categories of security would be interviewed and asked whether there are any issues that would qualify as human security threats in their region - threats that present a critical and pervasive vulnerability to the vital core. By way of illustration, for the environment, this could be an annual flood, or for personal, this could cover a high risk of landmines. There are no limits to the number of threats in any category as the only criteria is that they surpass the threshold of human security. The most important point about this stage of the methodology is that it has reduced a seemingly endless list of threats (anything that can serious harm an individual), down to only those that in practice affect a particular country or region. By shifting scales from the national to a local focus, human security becomes 3 4 Most global indexes use this 'laundry list approach'. a manageable concept, going from hundreds of threats, down to a handful. Fig. 3.1- Stage One Diagram: Threat Assessment Environ. Economic Political Personal Food • H Q Health The above diagram is a visualization of stage one. For each of the six human security subsystems, boxes a, b, and c represent threats chosen by the experts. The vertical lines separating the subsystems demonstrate the isolation of the raw data within disciplines. 3.2.2 Stage 2: Data Collection and Organization Now that the human security threats affecting a country have been determined and classified, data detailing them must be collected. These data can be both quantitative and qualitative, but all must have a spatial orientation. And in the spirit of the decentralization aspect of human security, they can be collected using local researchers, the NGO community, government ministries and International Organizations. A key to this stage of course is data availability. It is argued that the challenge is best addressed by looking at the sub- national level, by using disciplinary experts and by focusing only on relevant threats, those that surpass the human security threshold. A central criticism of broad measuring indexes is that it is incredibly difficult to get accurate data for all possible threats from all possible countries. The methodology used in this thesis addresses this legitimate concern by looking at where the data does in fact exist, at the sub national level. Even within a particular region, relevant data that might identify human security hotspots is not collected or compiled by one organization, it must be found from an array of local, disciplinary experts. This is because whereas there are many organizations thar collect very specific spatially referenced data sets, none collects 41 them a l l . For example, for a "vulnerability to flooding" data set, one could go to a river basin management NGO, or for violent crime statistics, one could go to the country's ministry of the interior. Neither, however, is likely to know anything about the other's area of expertise. Finally, data availability is often dictated by the presence and severity of the threat itself. Yet, this connection works in tandem with the human security threshold as if the problem is severe enough, the chances are very good that there are data detailing it. Due to limited resources, most data detailing human security type threats in a developing country will be based what are of greatest regional importance. There should, therefore, be significant information on the determined threats. Subsidiarity35 is particularly important to this collection process. Although information on all threats will not be available for all areas, the data are likely to cluster in the scale and regions in which it is relevant. For instance, if a flood affects only region x of a country, there will not be hydraulic data for region y. Once data sets detailing each threat are collected, they are organized in a GIS by their spatial reference. This reference can be either a political boundary, a coordinate or a grid space. What is important is that there is a link between threat severity and location, or, space. At this point, we can now determine the level of threat for any point in our study region for any of the initial threats. The notion that problems be addressed at the relevant or, appropriate scale. 42 Table 3.1-Stage 2 Table Health Political Province Name Threat A Threat B Threat C Threat A Threat B Threat C 1 2 3 4 5 A n Example of the Stage Two data set. Note that the spatial reference, in this hypothetical case province, does not have to be the same for all threat data sets. In the final table, however, they will all get rounded down to match the set with the finest resolution. 3.2.3 Stage 3: Data Visualization and Analysis The final stage is to analyze and map the spatially referenced threat data. For each of the determined threats, we now have data sets detailing the location and severity of the threat within the country. As all information is linked to a spatial reference (province, city, coordinate etc.) we can map each threat using a Geographic Information System. This process involves three steps: base map creation, hotspot analysis and correlation analysis. The first step in stage 3 is to create the base maps that will both display the varying severity of each human security threat and be used in the subsequent spatial analysis36. The second step is to link the data sets to a digital map via their spatial reference. This is done by matching the databases using the common codes given to spatial regions within a country. Once this is done, each threat can be mapped. These base maps are called layers and will be the foundation for the subsequent spatial analysis. This process will be described in much greater detail in the Cambodian case study, see Ch.4. 43 Although at this point each data set represents a particular threat, an inherent focus of this thesis is to examine where this threat is the most serious. This is done by classifying the data based on its natural breaks before it is mapped37. This process will produce a map for each threat showing where the threat severity is 'high', 'medium' or 'low'. Table 3.2 Stage 3, Table 1 Health Political Province Name Threat A Threat B Threat G Threat A Threat B Threat C 1 1 2 2 1 2 0 2 1 1 2 1 0 1 3 0 0 2 2 0 1 4 0 0 2 0 0 2 5 2 2 2 2 2 2 The above table is a hypothetical example of the threat rank given to each threat in each geographical region. Low=0, Med=l, High=2 Critics of the methodology thus far may point out that this adds an element of subjectivity that counters the core principle of the methodology (ie conceptually broad A N D empirically accurate), but this nominal scale, high, med, low is not as subjective as it may first appear. As these threats were already determined to be of high relevance (above the minimum human security threshold - see HS Commission, 2002), where they are highest is both critically important and reasonably easy to decipher. Simply knowing that flooding is a problem in a country means nothing unless we know where this problem is occurring and its varying levels of impact. There are many other ways that data can be classified. 44 Hotspots Hotspots are regions of aggregated human insecurities. They are places that experience multiple human security threats of a high level. Although a country as a whole may experience many different threats, these threats are often regional - different areas afflicted by different harms. In some locations, however, these threats overlap. Presumably, a person in a region suffering from 5 threats will be less secure than someone in a region with only 2 threats. Human security hotspot analysis is useful for a number of reasons. First, conceptually, hotspots demonstrate the utility of human security and the necessity of using a broad conception. They clearly show that people remain insecure while not at war (countering traditional security), and that within their border they are suffering from a wide range of possible threats. Second, aggregating varying data sets spatially facilitates a degree of interdisciplinary analysis that is rarely actualized in human security studies. By way of illustration, although many people know where flooding is harming people, and many people know where poverty is worst, few people know both. Also, difficulties of data aggregation and cross discipline communication often hinder well-meaning broad analysis. By limiting subjective decisions as to the spatial severity of various threats to the early discipline-specific first stage (threats assessment), the analysis avoids arbitrary or entirely subjective data aggregation. Third, hotspot analysis is practical as it aids significantly in development and humanitarian relief efforts. The practical benefits of knowing exactly what harms are effecting which region of a country are clearly evident. In addition, having all the information in a GIS system allows for easy access to vast mounts of data that generally does not get shared, let alone used widely within the development community. Hotspots are found by first separating only the regions with 'high' levels of insecurity in each of our threat severity maps. These maps will show where each of our threats is the most serious. 45 Table 3.3 Stage 3, Table 2 Health Political Hotspot Count Province Name Threat A Threat B Threat C Threat A Threat B Threat C 1 1 2 2 1 2 0 3 2 1 1 2 1 0 1 1 3 0 0 2 2 0 1 2 4 0 0 2 0 0 2 2 5 2 2 2 2 2 2 6 In Table 3.3 It is clear that province 5, with 5 high security threats is less secure than say province 2, with only one high security threat. A l l of these maps can then be overlaid 3 8 to show the regions subject to multiple high levels of human insecurity - how many 'high' rankings a spatial unit has received. This final map is dynamic in the sense that any data set used can be viewed by clicking on a threat. This will give practitioners the knowledge of exactly what threats afflict which regions of the country. Correlations The final stage of the methodology used in this thesis relates to one of the most cited criticism of the broad conception of human security. This criticism reasons that by lumping dozens of threats together in one category, human security does not allow for the separation of dependant and independent variables necessary for causal analysis. How are we supposed to decipher whether poverty causes violence if they are both called human insecurity, critics ask? I propose that grouping - not amalgamating - all possible threats under one heading, and subsequently into one data set, facilitates the very type of causal analysis critics often call for. For a description of 'overlaying' see Part IV: Cambodia Case Study or (Antenucci, 1991; Hey wood, 1998; and Bernhardsen, 1999) for an introduction to GIS. For a description of uses of GIS in the field of political science see Starr et al 1995. 46 Fig 3.2 - Causal Relationships Environ. — Economic Political Personal Food Health The above diagram demonstrates a grouping of a number of individual threats under the concept of human security and shows how variable can still remain independent as data sets and specific causes to allow for correlation analysis. This methodology facilitates direct causal analysis through spatial correlation. If there is a high incidence of poverty in provinces where there are high levels of violence, (and not vice versa), then one can be suspected of causing the other39. Spatial correlations are done by posing a series of simple logic questions using the GIS: Tf A and B then C (If high poverty and high violence the x). For example, using GIS one can create a new map showing only those regions with a high level of poverty threat A N D a high level of violence threat. Or, it could show only the areas with a high presence of landmines A N D a high level of poverty. By conducting a series of these binary questions, spatial correlations that would be very difficult, if not impossible, to see by manually comparing the very complex data sets, should be revealed40. Another final important feature of this methodology is that in order to put this information into context and to facilitate development operations, features can be added to any of the maps created. This could include political boundaries such as states or provinces, infrastructure such as roads, airports or cities, or topographical features such as terrain type, elevation or land use. An important question that could be addressed using this type of spatial analysis is 'under what conditions, i.e. poverty, is there likely to be violent outbursts? This could be done by using the methodology on historical data, an exercise that would add substantially to the credibility of the correlation. This could also be a more complex question such as if ' A and B and C but not D, then E ' . 47 3.3 Summary Three points are crucial to understanding the goal of this methodology. The first is the understanding that the methodology is intended to be used only after a country or region has been identified as having a human security risk. This identification would be done using a much broader, universal index such as the Human Development Index or the Human Security Index. Although these indexes provide valuable national data they do nothing to either locate or identify specific human security threats within the country. It is at this point that methodology would begin. Second, is the critical importance of the broad UNDP definition. This may be criticized for bringing too many variables into the system analysis, and in so doing, over-generalizing what should be a sensitive study. The definition of human security used, however, is intentionally open and responsive to a wide range of indicators. Once regionalized, the process focuses on serious concerns to individual security embodied in the chosen indicators, thus moving from an abstract to a specific definition of what threatens individual security in the particular country. To illustrate this point we need only look to examples of Bangladesh and North Korea. Most people would agree that individual security is threatened in both countries, but the source of this threat and the indicators used to measure it would not necessarily fall under the more specific definition of human security offered by the Canadian government. This definition focuses solely on the threat to the individual from direct violence, which is not a significant concern in either of these countries. In North Korea the threat comes from food security; in Bangladesh the threat stems from environmental security. Both of these insecurities would be recognized using this methodology. I believe that this methodology has the potential to: • Facilitate collaboration and information sharing. • Aid coordination in peace-building and peacemaking initiatives. • Provide a better overall picture of a vulnerable country. • Help compare and validate multiple data sources. 4 8 The mechanisms of traditional state security are clearly not inclusive enough to deal with the vast array of potential threats to individuals. This inadequacy as led to the evolution of the concept of human security. By shifting the referent object from the state to the individual, human security has refined its resolution and hence improved it capacity to deal with these new challenges. However the theoretical shift, although necessary, is not sufficient. We must also develop a way of identifying and locating these insecurities. I propose that a GIS based methodology could fill this vital gap between the theory and the practice of a human security based foreign policy. 49 Chapter 4 Cambodia Case Study As the core principle of this methodology is contentious - that a measurement of human security can be both contextually broad and analytically specific - a case study is necessary to test its feasibility. This was carried out through a Cambodia field study, from November 2002 to January 2003. I was based at the Cambodian Development Research Institute (CDRI) in Phnom Penh, and the first two stages of the methodology (threat assessment and data collection and organization) were successfully conducted in full. The final data analysis stage was completed at U B C from November to August 2003 and the results will be returned to all NGOs, IOs and Government Ministries that provided data. It is hoped that the final tool will be used by practitioners and academics to assess threat levels and facilitate development and humanitarian relief efforts. There was decisive support among development researchers and practitioners in Cambodia for both the concept of human security as well as for the methodology and the tool of GIS analysis. 4 . 1 Why Cambodia? Three criteria were used in selecting an appropriate case study for this methodology: level of insecurity, data availability and feasibility. First and foremost, the methodology needed to be tested in a region of high human insecurity. Cambodia ranks very low on almost all indexes - 130 th in GDP Per Capita, 130th on the HDI, 140 t h on the HPI. 4 1 Second, there must be sufficient data to implement the methodology. It is possible that in the most insecure countries, such as in sub Saharan Africa, there may not yet be enough accurate data. The process becomes more difficult and resource intensive the less data there are and this factor was particularly important for a single researcher with minimal funding. Due to a significant UN presence since the fall of the Khmer 4 1 This is in fact a core aspect of the methodology itself, that we already know what are the insecure countries. What we need to know is why these countries are insecure and where the problems are worst. 50 Rouge there has been much research and data collection done in Cambodia. There are over 400 NGOs operating in Phnom Penh alone. Finally, linked to the second criteria, the implementation must be feasible. Data must be accessible, the region must be sufficiently safe to work in (not in serious conflict) and aid of a host institution in preferable. Cambodia is safe to live and travel in, major conflict ended in 1999 and the Cambodian Development Research Institute offered to house and facilitate the project. 4.2 Cambodia Background What oppresses me, more still than the closed eyes of the dead who fill the sandy paddy fields, is the way the West applauded the Khmer Rouge, hailing their victory over their brothers in 1975. The ovation was so frenzied as to drown out the protracted wailing of the millions being massacred. (Bizot: 4) After 1968 Cambodia was subject to unparalleled misery at the whim of repeated devastating irony. This destitution came from both the hands of foreigners, 115,000 illegal US bombings, and from within, with the rise, rule and genocide of Pol Pot and the Khmer Rouge. It was ignored by the world and then flooded with foreign assistance; throughout both, people suffered42. The fighting has formally stopped, the international community has been active in the country for ten years, and the Khmer Rouge will be subject to War Crimes tribunals in the coming months. What is left is only a shell of what Cambodia used to be. Cambodians are still not safe, they are subject to multiple threats and they live in dire poverty43. Cambodia arguably reached its cultural, artistic, political and regional power zenith during Angkor period of Khmer history from the 9 t h to the 15 t h Century. Since then, however, this small, historically prescient, and once confident country, lived in relative peace as part of the Union of Indochina and under the French Authority up until the mid-20 t h Century. Ironically, having finally achieved full independence and trying to stay out of the conflict in Vietnam, Cambodia's recent history took a dire turn in the mid 1960s under the The exact dates and sequence of events for the following historical overview were taken from the Canadian Foreign Service Institute's 'Cambodia, an Introduction'. 4 3 For arguably the definitive modern history of Cambodia, see Chandler, 2000 51 rule of Prince Sihanouk, with the conception of the Khmer rouge and at the mercy of American and South Vietnamese policy. In 1968, the Khmer Rouge formally began their armed insurgence, and in 1969 the US began an intense four year illegal bombing campaign of North Vietnamese outposts in Cambodia. In 1970 the US backed South Vietnamese moved in to take over the country. In the chaos, Sihanouk was overthrown by Lon Nol, popularly seen as an American Imperialist, and in 1975, the Khmer Rouge took control of the country. What followed was a genocide comparable to that in Algeria, Nazi Germany, and Rwanda. As an inconceivably literal and brutal manifestation of a Maoist agrarian revolution (Pol Pot spent significant time studying in China), all cities were evacuated (ironically, under the threat of further US bombing) and all educated people were killed: teachers, doctors, professors, politicians, writers, artists. The numbers reported killed in the revolution (genocide) vary from 800,000, if solely those murdered in camps are counted, to over two million if indirect deaths from starvation, overwork and forced migration are rightly included. In 1979, the Vietnamese took over Cambodia, ending the genocide and installing a Vietnamese styled peoples' republic. A coalition was established between Khmer Rouge defectors and the old political elite that had survived by fleeing the country. In another bitter irony, the Cambodians were saved by their historic enemies. The Khmer rouge then retreated to the north of the country (see 'hotspot analysis' on page # for the insecurity levels of this region), and launched a guerrilla campaign that lasted almost 20 years. In 1990, the Paris Peace Conference for Cambodia establisheda broad plan that lead to a ceasefire, a peace accord, and an interim authority lead by Sihanouk, all overseen by the largest peacekeeping operation in history, The United Nations Transitional Authority in Cambodia (UNTAC). In 1993 U N T A C oversaw the first democratic election. Although FUNCIPEC (the royalist Party) won a majority, Hun Sen and the CCP (the Vietnamese backed Communist Party) had control of the military, resulting in a coalition government, with an awkward arrangement of a first and second prime minister. This tenuous situation was exacerbated by the withdrawal of U N T A C and its inability to wipe out the Khmer Rouge, who continued guerilla 52 attacks until the death of Pol Pot in 1998. In 1997, Hun Sen lead a successful coup (although many in Cambodia do not recognize it as such) over his rival prime minister and won the election in 1998 to become sole leader. Since then the country has experienced relative peace, with the absence of full blown civil conflict and a steady decline in violence and crime. This being said, Cambodians are not safe. Corruption is rife and people are dying at alarming rates. Although Cambodia is now 'traditionally' secure, its human insecurity is having a devastating impact on the people. With the upcoming election and fears of violence certain to garner the spotlight, optimistic conjecture is looking particularity naive. It is difficult to conceive who will address or even recognize all that is truly affecting individuals in Cambodia. An independent analysis of human security threats is therefore both feasible and timely. 53 4.3 Spatial Contextualization Figures 4.1 to 4.5 show Cambodian in various scales and with varying political boundaries. Fig. 4.1 Map of South East Asia 4 4 Base map data were provided by Geospatial International, the C M A A and C1DA. Political Boundaries Used in Analysis Fig. 4.3 Maps of Provinces and Regional Districts 24 Provinces 185 Regional Districts 56 Fig. 4.4 Maps of Communes and Villages 1628 Communes 57 4.4 Implementation of Methodology The methodology that follows moves through the same stages covered in Ch. 2. 4.4.1 Stage One: Threat Assessment4 5 The discussion in Chapter 2 indicated that there is yet any consensus as to a common set of variables covered by the human security paradigm, consequently, a manageable list of threats afflicting Cambodia must first be determined. This was done by listing all possible threats that surpassed the threshold inherent in the human security definition4 6, in each of the six categories of security (economic, health, environmental, personal, political and food): viz. those that represent a critical and pervasive threat to the vital core. As stated in Chapter 3, identification of 'key threats' would ideally be done by a panel of experts from each of the six categories. As there were limited resources for this case study, a modified three step process was used to determine the applicable threats. First, a literature review was conducted using the library of the Cambodian Development Research Institute This library houses every significant report written on Cambodia in the last decade, an invaluable resource for this project and for Cambodian research in general. The literature review was carried out by searching each of the categories of human security for reports and articles. A database was created and all relevant reports collected. While this review did not reveal the exact final list of threats with any certainty, as local expert were required for this, it did provide an invaluable background to the Cambodian situation that was essential to the success of the expert consultation and interview process that followed. For a chart summarizing stages one and two, see Appendix 3 4 6 Human security is the protection of the vital core of all human lives from critical and pervasive threats. Individuals require protection from environmental, economic, food, health, personal and political threats 58 Second, expert consultations to determine key threats were conducted using in-house researchers at CDRI. As CDRI is a well respected, multi-disciplinary research institute, it provided an ideal starting point into the often insular Cambodian research and NGO community. Each of eight interviewees, all senior CDRI researchers, were given a two page letter outlining the concept of human security and the methodology (Appendix 4). They were also given a table to fill out, that would direct me to experts in each of the six security categories (Appendix IV). These experts came from the NGO, IO, Government Ministry or research community. What was important was that they had a firm understanding of their area of expertise, their particular category of security. From these in-house consultations, an invaluable list of 40 interviewees was compiled. Perhaps most important was the fact that I could use the CDRI researchers as entry points to future official interviews. As they often recommended senior NGO officials and ministers, this personalization was absolutely critical to the success of the case study. Third, following these initial stages making contact with key informants, a final series of 65 interviews were conducted. A data base was then created to record the details of the interviews that were conducted. Table 4.1 Example of Interview Database in Secur ity Area NaniL Contact Info Organization Reforre dB> Interview Status N ites 5 Food Slk Boreak 212137/8 ufp-pnompenh@wf p o r g WFP Sopheak Wednesday 27th afternoon Interesting revelations. They are doing very similar things, although directed specifically for the a id of food distribution. Their poverty map will, for my analysis, constitute the economic and food data set. Poverty Is defined by consumption expenditure... 6 Politico 1 Koul Paulla 012 942 017/012 942 019/023 884 150 COMFREL (free and fair elections advocacy) Sopheak wed nov 27th 3:00 Eva warned of possible bias and advocates against local authority abuses Wants me to do the mapping for them. Should get a good data set of election results and election Irregularities. It has been suggested that areas of hlnh onmnetltlon Source: Field notes 59 As will be discussed below, due to time constraints, these interviews served a dual purpose. With respect to this stage of the process, however, they served to determine (based on the experts' own subjective perception) exactly what threats surpassed the human security threshold in each category. Again, this was essential in reducing what could be hundreds of possible harms to a meaningful and measurable list of threats. Subsequently, the following list of human security threats for Cambodia was developed. Fig. 4.5 Cambodian Threats in Security Category Environ. Floods Droughts Economic Poverty Political Corruption Human Rights Abuses Personal Landmines U X O ' s Violence Robbery/Theft Small Arms Food Starvation Malnutrition Health HIV/AIDS Malaria TB Dengue Fever These threats, organized in disciplinary isolation, constitute the critical and pervasive threats to the vital core of Cambodians. 4.4.2 Stage 2: Data Collection and Organization Now that the critical set of threats had been determined and the broad conception of human security had become manageable, data detailing these threats were collected. As the data was inputted into a Geographic Information System for analysis, the only criteria for this data was that they should have a spatial reference. In the case of Cambodia, this could mean that the data was disaggregated politically, to the village, commune, regional district, province, or by a coordinate system, a point on a map. It should be noted that at this point, the data was still in its raw form. That is, the unit of analysis is disciplinary, and therefore may be difficult for non-experts to understand. For example, flood modeling data are of little use to a rural poverty specialist, and vice versa. In addition to collecting data detailing the determined human security threats, all available GIS data were also collected. For reasons described in the GIS section, a number of base map data sets required to build a GIS database. These were collected by interviewing most of the GIS experts in Phnom Penh. 60 For this case study, the data were collected through the interview process outlined above. The two stages, defining critical threat assessments and data collection were done in an overlapping fashion due to time and financial constraints. The snowball effect of interviews leading to suggestions on finding data turned out to be very successful. Although there are more that 400 NGOs in Cambodia, I very quickly developed an appreciation for their particular biases, relative respectability and key people or organizations collecting . data. Data Quality Although there is a significant amount of data available in Cambodia, much of it is of questionable reliability. This is due to a combination of political bias and dubious collection methods. Cambodia is a country with a deep rooted political divide. Supporters of the ruling party, the CCP, often do so with impunity, wholly supporting the often corrupt, illegal and undemocratic practices of the government. The data that these individuals release, no doubt lean toward a government position. Supporters of the opposition parties, particularly the Sam Ramsey Party, the harshest critic of the CCP, are often equally entrenched in their bias. As mentioned, this includes many human rights NGO's whose primary objective is to rightly point out government human rights abuses. For some NGO's , however, this clouds judgment regarding collection methods and data analysis. This being said, after I spoke to many informants in the country, these biases were revealed and the relative reliability of data sets became very clear. This demonstrates how at this point, a degree of subjectivity is healthy to the assessment. Not all data sets are accurate and reliable, but this does not mean that all of the sub national data are corrupt and unusable. It is precisely this mentality that leads to a discarding of all data but the vague national level indicators produced by International Organizations. The fact is that much of the data are collected with integrity and reasonable accuracy. The few bad examples should not discredit entirety. 61 An example of a subjective decision being made on data reliability is the case of the Partnership for Local Governance data set on village level insecurities which includes levels of thefts, conflicts over land and violence in the home. They collected a detailed survey at the village level, an incredible undertaking considering there are 13,000 villages in Cambodia. From much discussion about the data, I learned that the first round of collection was highly suspect because the village chiefs did not know how the data were going to be used and so were worried about highly sensitive security information being used against them. However, the primary purpose of the data was to return it to the villages so they could compare themselves to other villages in other parts of the country47. Because this first round proved to be helpful to the villages, the second round of data collection is seen to be highly reliable. After hearing about the first round's reliability, I decided not to use this particular data set. However, having later received the second round, it is included in the analysis. Another example of data reliability concerns the human rights NGOs. While many collect information on human rights abuses, some do so with significant bias against the government. Although this bias is arguably warranted, it invariably affects the collection methods. Of the two most respected human rights abuse data sets, one N G O would not release them to me due to confidentiality concerns, yet the other was released so is used in the analysis. Again, the indicator could have simply been excluded all together, but with research and a degree of subjectivity in the data selection process, meaningful information can be attained. Data Collection Results Significant amounts of sub national human security data were collected. Table 4.2 lists the data relevant to the 6 threat categories, 13 identified Cambodia specific threats, the spatial coverage of the data, the time of data collected and the original source. The data is organized into human security categories, other statistical socio-economic surveys and general GIS information. This is one of largest existing Cambodian databases. 4 7 A package was given to each village, including a map of the village and the census data, both created 6 years ago by the national census and never disseminated. Charts were also included showing comparisons with other similar villages. 62 Source I WFP | National TB Program, Mnistry of Health | E (0 & t 0 8 LL 5 « c .S I National DHF Control Program | National Malaria Control Program | National Malaria Control Program | Hellen Keller | Mnistry of water resources | mrc I mrc I mrc I mrc 1 Q a Malaria control program | Geospatial 1 Geospatial, OA I Licadho 1 Licadho I demographic Health Survey | 0 B! EC Small Arms project ACRA | EC Small Arms project ACRA | COMFREL I UNHCHR I USGS I 2 Geospatial I Geospatial I JICA I Mnistry of health. Malaria control pogram j Mnistry of Ranning I Mnistry of Ranning I Mnistry of Ranning I Time 2002 2000, 2001 ! 1999, 2000, 2001, i 2.0021 2002J 20011 20011 20001 20001 ten year 20001 20001 20021 20021 20011 20001 present I 70-79 | 2001, 2002 I 20021 20001 2001 | 20021 20021 20021 20021 20021 19601 20021 20011 20011 20021 20021 19981 20001 19991 Coverage Dstrlct OD province OD | OD I astrlct | V 0 c > 0 ct pixel | polygon | pixel | pixel | pixel province province village 1 polygons I village I village I province I province j comrune | village I 7 provinces | 15 districts | Proctnce I Dstrlct | Commune | whole country I 2/3 of country j whole country I whole country | w hole country | w hole country | village I village j village I Data | Poverty Map (8 maps) | Annual Statistics of TB j Dengue Rates j Dengue Map j Malaria Rates I Malaria Maps (5) I Enerria Rates I radarsat image (highest water level) I 2000 modeled flood levels | average w ater levels I 2000 flood levels I 2000 flood duration | Number of people affected 2002 flood 1 Nurrber of people affected 2002 draught I Conflicts over land I Land Use I Level One Analysis I Victims Database 1 Level one and Bombing locations 1 Politically motivated violence I Reported domestic violence, 7 types | rates of domestic violence j Reported insecuries (rape, murder, banditry, robery) j Small arms survey (33 questions) | high security risk due to small arms | Electoral Irregularities I Sectoral Violence I Bection Results I 1:50,000 Digrtal Georefenced Topographic | 1:100,000 Georeferenced Topographic (hard and soft)] Level one I 0 n! georeferenced I roads, water (digrtal georeferenced) | roads (Hard) I location I 4 eds I •o o 1cd I Threat | >. <L) > & Starvation | P HIV AIDS | Dengue Fever | Malaria I (0 E Rooding I Draught 1 Land conflict 1 Landmines 1 w 0 3 Human rights viotations I Domestic Violence I Incedence of insecurities j Small Arms (Data Pending) I Democratic Hinderances I MAPS I Gazzeteer j Boundaries and village points I Inf astructure I Operational Health Districts • I 1998 Census j 2000 Demographic Hearth Survey I 1999 fishing Survey I 1996 Socio Economic Survey j various pdf maps I 40 reports, curoborating each of the 6 categories j Category | Economic | V 0 £ Health Environment j Personal j Political I Geographic Data Surveys Reports | o 0 0 Data Organization The next stage was to organize the data detailing each threat by its spatial unit, eg province, district commune or village, so that it could be mapped and analyzed in the GIS. A table was therefore created for each threat showing solely its spatial unit, the reference code of that spatial unit and the data set. For example, the base table for a data set (infant mortality per 100,000) using provinces as its spatial reference will look like this: Table 4.3 Example of Data Table KHETCODE1 PROVINCE Infant mortality per 1000 births 1 Banteay Meanchey 78.2 2 Battambang 98.0 3 Kampong Cham 107.9 4 Kampong Chhnang 129.3 5 Kampong Speu 68.3 6 Kampong Thorn 64.5 7 Kam pot 100.4 8 Kandal 89.2 9 Koh Kong 70.7 10 Kratie 71.3 11 Mondul Kiri 169.8 12 Phnom Penh 37.6 13 Preah Vihear 71.3 14 Prey Veng 111.0 15 Purs at 139.4 16 Ratanak Kiri 169.8 17 Siemreap 77.9 18 Sihanoukville 100.4 19 Stueng Treng 71.3 20 Svay Rieng 102.0 21 Takeo 96.0 22 Otdar Meanchey 77.9 23 Kep 100.4 24 Pailin 98.0 99 Tonle Sap Source: Demographic Health Survey, 2000. However, as the spatial unit became more refined, eg from the 24 provinces down to the 13, 336 villages, the data sets became much more complex. For instance, if the commune level was used, there was 1629 records as opposed to 25, one for each commune. Similarly, if simply a coordinate was used, then there 64 was a record for each data point. For the US bombing data set for example, there are 115,000 records, one for each bombing sortie (see Table 4.4). 65 S 3 gg g 1 g si§ § j . § §,§ §§§.§§§!§ §. io!*?! I ;3iS;8i3:~ Jlll i l l l l ifs | l | |» l * | * | | | | | * | | l g g s S_S.s s S11 § 1 2 2 !2I2 a.;o.! !Q.;O..V_ 0 ; 0 ; i O i O : ^ g ; g -!l!SSI|ll!IS <*t-:a,:a-;a-rS , Ifll igllffii 'Sla^Sil i ' iSj l^ljJpJIIIIMI: 3333333 yil l i i l l i l ! ; !ri^|2i2!2: ! Ui2 ; I is^i 5 3 i _j<up;3:0:as i3=S;0;?!sJ3:5f= ^ 51 g S gfi S 3 § 8 S1 Sl8,l S;5, !§:§: :(D;<D; i SI 3ia.ia.i--; :C k p i l l i i l i ^ l l ^ i l l 191 |9|9|9|9| |9 |9js!9!9l| | I f l ! l ? 1 1 1 I f | 9 nftliiaJaJaJaJ!£ia.!aJa.iaJa. ifi ail: ilia :o!8 j 8 S SI ,ligllll|ill|NiilMilft^!S!l[lililSis I !9J I I9J9I9I3 » S i §IS|i!i!S!lS!° i Hia. io.la.io.ia.io.ia.in IA ;5i3l5!Ei3l8i8i3i i :!: ai l ia I s la III Ii «:<r:i£:| Hi li II • ? 3 2 t | 5 S 8 8 i i?i i I I —4—;—I— It: IS; j"S! .:U.:U.;u.;u.ju. , 8 . 8 8 . M i l PPPI |o|o| I I i i k i r :Ll6l8i3|?!?iLI-ti-fia!aUlviT!T: : t f : • * : r n - "3 ju-.;u. \u. \< iu. lu. :m ita in. [< • <'•<• j O i O i O i O i O l O i O IHI li: j ? 3 3 l 1131318! isBI * a ip l j i l i : l i i l i e i : l c | i i 9 i S ° o ; o i O i O i o i c li|t|ililllilllj|ll 1l|a|g]||gMS!p!g|-|Si!5|S!«i|§|s|g| 3 " 3 : i 1 M I I <U ' I S H I R ' M 3 I ca Q 00 c e o m D -a ' ; l j I l I ; I ; I i Q : O i Q ; ) : O j O ; O i O : o i O ; 0 ; O i O ; 0 : o : 0 ! 0 ! O i O j O j o ; o : b i t !i5lftl5ISISiliiil§iiSI?l?l; J;ffi;ID:«>:IO:CO:tD:<D'<O-lO;iO;cO:35;2j[0 i O I O I o ;o ;"o EO;O j o i o ' i o i l l l l l l l 5 a s a i O : O i O i O i O i O T C y : 0 J : C ! B 6 8 ri S i5S S S ffh S ? ^  S 9 S ftsi •; ? k g 8 § g S s * 3 J s si' s a s g a .•» a H ^ k s s 1M § § S l i l i i i ^ ^ l i i i l i i i i i i l;8!ii i|5ls j | 0 ; O j o i ^ g a S S ^ | s s:y s s.s I II o ! S X H S s 3|S|IS|i§ISI&i6|!§|§:g|g|giglS|gi^ tZ> 8:5ialaMis|s;ais!a|8i5l3!3|s^ I l i M I i ; M M M M i i I • ; M M M M ; : j j i I M i Ij3 U . . I . . . U . ! L . . U . J . . ! [ [ . J . : . . ^ ! . . \ u J . J . J J . i . j . . i j . J . ! J J . . l J I..I..IH At this point in the analysis, I generated a data table for each of the human security threats linking the data to a particular spatial unit4 8. It was then possible to create a GIS data base, to map each threat and to perform the hotspot and correlation analysis. Tables were created for the following human security threats: Economic and Food Security4 9 Poverty- Percent of population below the poverty line (commune)' Health Security Dengue Fever - new cases per 100,000 (Province) Malaria - % of population at risk (Province) HTV/AIDS- % of infected sex workers (Province)5 0 T B - Cases per 100000 (District) Infant Mortality per 100,000- (Province)5 1 Personal Security Landmines - Percent of contaminated villages (district) Small Arms-Percent of physical injuries due to a gun (Province) Domestic Violence-Percent of Married Women Experiencing Domestic Violence (Province) Violence- conflicts over land, theft and robbery, and violence in home (Commune, aggregated up from village) Political Security5 2 Human rights violations per 100,000 (Province) Environmental Security5 3 Flood - Affected People per 100,000 (District)5 4 Drought - Affected people per 100,000 (District) See Appendix 5 for all data tables. 4 9 After much discussion, food and economic security were combined as one with a poverty indicator. This method is advocated by the World Food Program who believes that when a country is in a low enough level of poverty, income means nothing. The only measurable expenditure is how much they consume. This is measured in Calorie Intake. 5 0 Data are available for other high risk groups but not for the entire population. I was told that number of sex workers was a particularly good indicator for HIV/AIDS incidence, especially in Cambodia. 5 1 Although not listed as a specific human security threat, I was told by many that it was the best single indicator of health in Cambodia: 5 2 It is possible to get data on election irregularities. However, duo to time restraints, this was not collected. 5 3 Data on deforestation and particularly land concessions is incredibly controversial at the present time. Although significant data exists, including maps of future logging areas, it was impossible to acquire. 5 4 This data is collected by the ICRC. It encompasses all of their humanitarian assistance and details exactly who received what and where. I have also acquired radarsat images of the 2000 and the 2001 floods which are used for the poverty overlay analysis. They were not used for this because they don't indicated necessary socio-economic impact. 68 4.4.3 Stage 3: Data Visualization and Analysis Data Visualization Now that all of the data were collected and organized by its spatial reference it could be inputted into a Geographic Information System. Once base maps were created, a common scale of threat level was assigned to each data set. They could then be overlaid, all together, to find 'hotspots' of aggregated human security threat; and, the threat scores of various hotspots arranged in binary pairs, to examine spatial correlations. Finally, base map data were added to help contextualize the maps. Spatial Joining In order for the data tables to be mapped, they must be joined with the data tables of the shapefile55 representing the map boundaries. These tables are joined by a like field. In this case, the like field is the GIS code given to the spatial unit5 6. This allows for any of the information in the threat table to be mapped. For example, by joining the Domestic Violence table with the Provincial Boundary table by their like GIS codes, a map showing Domestic Violence Levels can be mapped57. 3 3 Shapefiles are the complete package of files, one of which is a data table, needed to display information in ARCview GIS, the program used for this analysis. 5 6 Matching these codes caused significant problems with the Cambodia data. As Cambodian researchers have just begun to use GIS data, their coding system has not been standardized. Also, at the commune and village levels, there is considerable inconsistency in the names of political boundaries. This discrepancy in codes was overcome by manually matching all codes in an excel spreadsheet then importing them back into the original data sets. 5 7 For the base level maps, all data was organized according to its natural breaks. This means that the data is categorized into, in this case 5 divisions, based on the way it clusters in the set. 69 Fig. 4.7 Diagram of Spatial Joining Process (next page). GIS CO« -MOVHCE Province Boundary Shapefile • I IMHT "aamaiay HAHITAY H l m j f i • o*a*nt>»o tnmninriiia I *mng Own .*U(«WICJWJ ' «-T»n9 CHranj - I U I - M , . i .«,.,„; T>™n IWMHa THOU i*«kMCXM'/tJ.£ .IT. 'J»4Jtl*fi« 701 • n w M N /it i uwrtt?no.:2ti MM] 1K7 v i rn* no UO'*) MO Joined with Domestic Violence data table 05 '06 as cooe KHetcooe t PROVNCE 1 Banleay Maancnay 0 2 2 Ballambang, 3 Kampong Chan 4 Kampong Chhnong 3 Kampong Spou 8 Kampong Thorn 7 Kampol « 8 K o^da! 0 9 9 Kofi Kong " 10 Knlia , n II MondUKM l2PhnomP»nh O 13 Praah Vihaar , 1 4 M Pray Vang 1 5 ISPUPMM , ' 6 16 RaWmak KM " 17 SMmraap 18 SunnoukWla ,'9 -9 Sliiang Trang 2 0 «"0 S«y Riang •2' 21 Taaao 22 22 OUiaf Mtjanclwy ^ 23K.p 5 9 M Torte San Results in Combined Table (linked by GIS Code) <"ROVINCe is jf*i» jy Maoncrury 4 Kampong I Jiniwig 5 Kanpong Spay a Kampong Thorn 12 PMgM Pnnn 14 Pnty Vang PROVINCE 1 i lAlfTFAY MFANCHEY HA n AMOAMG <AMPOMG CHAM KAMPOHG CMMNANG KAMPOMG SPEU KAtOM, <OM KONG KfUl« '•••1*1 MHI niNCM pi-;m PREAH V* M , , ( FHffY V/gNG •<ATANAK KIR] SIHAN04JKVLLE STUcNG TR6NG SVAY RIEMG IAKC0 < I UTAH MEANCHEY KEP M0S42&4/, l.'JWJ 11 SOT 12sse; ran M8087066a 17; •.l!<J7H.,*4 Jfifl •5&90710DS.M1 12441423000.7:18 1383801147 1'W '211810:1311 488 12230793153 7U1 1 I M I 1 j.i 708 17 IH4U Prt BM 1 Hiwrams /«a M7i;o«;ij887T) 11721188608.131 i402iaiisa 137 120787O94S7 058 • I ID 14620291*3.703 4-81036370 084 <if?6ersi.43a 1102702/10 126 2503S5Si")7 iB2 PERIMETER G*S COOE '.336*4 AM Dl <~*M4/9 700 02 '73974 754 D3 129003 107 fw »03763 3te 03 '"fx I 087 1)6 •09329 778 07 J7788 410 t)8 W0782 309 t-9 'B87T2SM '10 <*8432 JU8'l 1 114924.041 '12 '11810.803 '13 G838.191"H 0278S 480"IB U8403 1)00 ?n 141902.087 1\ -d660S DOS '22 110001 U87 171511.008 *M 371343.1184 '<9 (•Hfjvwce % oi Vomon Atjuiad lljrxwy I.tcanttwy 12.3 •|j(;..(li(,» : ; 8 9 Karri iiong Cliam I8.B ilampong Chnnang 198 Kampong Speu 2.T 13.41 Kianaw 11.1 Karwd 17.9 Koh Kong 200 K/mita 13.9 9.5 P*™m Perm 17.2 fmari Vihaw 13.9 Pray Vang 7.8 •*J(.wuk - L.1 •1 s Swrwaap 27.8 SawnmAJfle 11 1 %*u*ng imng 3.9 J*»Y R*ang in a 13.2 fUdiB Maanchay 27 • M 7 0 This process of spatial joining was undertaken with each data table and their corresponding spatial representation boundary shapefile: for instance, province data were matched with province boundary shapefile, regional district data with regional district boundary shapefile and so on (See Appendix 6 for resulting base maps) Creation of a Common Comparable Scale Although each of the threats could now be mapped independently, as they all use different units of analysis, i.e., per 100,000, percent of contaminated land and so on. Consequently they could not be combined or aggregated into one table and subsequently one overall map showing 'hotspot' locations. In order to do this a common comparable scale must be created. For this analysis, the common scale was a simple ordinal ranking of high, medium and low threat for any given spatial unit. These three levels of threat, represented by a 2 , 1 or 0 respectively, were determined by classifying the data set in the GIS by its Natural Breaks 5 8. This grouped the categories based on clusters of data. Critics of this method will point out that this is simply using the same type of subjective scale that is criticized in methodological comparisons, particularly at global level, of broad indices. There is, however, a fundamental difference. The indicators used in the Cambodia case study have already been generated through a subjective analysis; in other words, they have already been deemed human security threats by the Cambodian experts in stage one of the methodology. Subsequently, sorting the data by a nominal ranking is simply highlighting where the threat is the worst. With respect to the Global Environmental Change and Human Security (GECHS) index, for example, a similar nominal scale is created, except there is no indication as to whether a particular threat affects the country in question. If it does not, then the nominal scale is useless, as high, medium and low mean little if the indicator does not represent a legitimate threat to begin with. Natural Breaks in one of many options for classifying data sets in Arc View GIS. It was chosen because it most accurately captures the bulk of the data without complex statistical reasoning. Other options for classifying include Quintiles, Equal Intervals and Standard Deviation. 71 A n additional step used for this analysis was to isolate all of the data with a threat level of 2, or 'High'. The resulting 13 maps (See Appendix 6) showing only where each threat is high, will be used for the hotspot and correlations analysis. It is conceivable, and perhaps preferable, that this process be carried out using an international standard of high threat for each category, rather than using the highest numbers in the threat data set. Experts could be interviewed, perhaps outside of Cambodia, and appropriate thresholds that were acclaimed at the international rather than merely Cambodian level, determined. This would ensure that certain regions would not get left out of a high threat ranking for simply having a rating slightly lower than the highest grouping. It would also, however, add another subjective decision, something the methodology attempts to limit. Below are examples tables and maps before the creation of a common scale, with 0,1,2 rankings and of isolated highs: 72 Fig 4.8 Diagram of the Creation of the High Threat Map for the poverty indicator Percent of Pop. Below Poverty Line | | AJI_dalaJlnat.shp Poverty Threat (High, Low, None) High Poverty Threat The final step before further analysis of hotspot location can be conducted was to merge all data, for all levels of analysis, (province, regional district and commune), into one final table referenced to the lowest spatial unit, in this case the commune level. This was done by first grouping all data, original data type and ranking by like spatial units. This resulted in three tables: province final, district final and commune final. These tables were then combined using the Union function. This function essentially overlays the maps (and therefore the data tables as well) putting all that was inside the more refined level (commune) into the coarser ones (districts and provinces). The final product was a table and map with all data listed by province, district and commune name and codes. 74 Province Level SCoeto^wnvwce j» 1! BanlAay MMnctWy 'T'''' IQJStutng Tnng 24*P*Nn X...XXXX is:Puri« '"!"" OiKohKong XIX 6iKampong Spau TiK«np« 2 3 : » C a i p " " ? " " 21:Tah«o 20:Sv«y Rtong • KmW wTpjiiy'viiiii i«1'«Wni*Kirt i i i M o S u K i l i " * _Wjtftjiiii._ F MAfWI IDOMV RANK \ 20t 2 13: ...... 12.6: '2] ""i2.3'[ " li "3331 2 34; .... XX'/iX' 13.6; 2! 27 Ml 2i "sorT 2]' """0.0* "o; 27*1 2 Sis: 2' ii; 12.6: 2: ' 13.9: i; a! ji X Ml ~ "it lis; 2* 13.Bi 1 ' " » ! " " " i| el""' Z"it 6j! li is' 6': "«fiT 2 96~ l'{ 'a'-'i] b" '"iei i- 7! it •.a: 38.9: 2i " "4iT "v. 66 "ii; 'if 20.6;"' 2 """at ii to: 1; 15.1: 2: 2.7] 14: ii » 1; a.4t l] 11.1] 0' 2 S3 2;""" *A* 'i • T'i.i; ""SI ZZZZZZZZ. 0; b"6'| 0; 13.2 ii " """it ..... .^ r 0: 7'.i\ "i'| i'iol 1: ne X X J T X " ' . 0] Oi 7"ii"' i; 17.8! ii a: Of 0.0; ol '""'"""7";a| bi "azat a* ""04: 2. 0.0T XXsXX a.si b' oi XX.CX'.'.'. 'nl]'"" 2 0.0T i".6? a ib; ii zzz'**.' ii""' 12.6: 2: 13 01 il u: X X ' X X 1]"""" "i'Sj i'i ieS] """ij "' 73i 0;""" """bi ~ZZ^-?T" XX? XX I7.2i V 401 11 'a.4- UAl „, "MX..... xxixx. tt'. ij o'Si* ij ixi": ij . . <?i ?; _. "til it 233i XX?C «.ai "i: Union [GtSCoda ipfSTWCT .. lOSipu.ChBjv ICMJPjiMh Nmti P M h io3:piir(Um Sroii " i"dz; Uooflkoi' ... .22061 Tiy^PiiMj, ""i^EGhiiJ'Krt .2202] Bwiiaoy. Amp*..... '"i714:5iuin "\ ""lyfj'isiv'iiu'"" 1709;f>tuat Bak'ong ' "l707;Puok 7_.i^iK^^"!!7."!V."7. ITOliearriuy Sf*l ...17TOiif«^«.Thorn 1701: Angkor Chum District Level !^.o?990...i^_^.r«.... .,,.0.00: 9-obT" "o.ooi" ,'J£obI Commune Level ....ft-pO; :QIS Code I i T O 3 0 2 i i1903051 '130301 190301] 1303021 :130304' f"""1363631 ' 130267' 220302! 1302031 i1302051 ] 1 0 7 0 6 130209: 130202 iwi'64' 10704" .' 136402' ;"" 171*206" 160909! 10705': i30403" 16763; ; 1 0 6 0 4 : 19C O D M M U N E Thma Kaev ICtioam Ksani iProaofc Moas TueJt Kratam iFUjmdaoh Sroo iPring Trttjm IKampong Sraiau Muoy ICrrangkaj 'Sjiu^iio'M!^ iKumru iKampong SmJau l-u ;OhhaobPl( irfma P'uoit iKukuMsn Choimg Sioong Spaan iPtinum Koic Kook KAktlxm :phum Tnmit [?i*iy! c ^ X X X ZZZ.". iPOV.FlANK ]PLG„TOTAL IPLG_RANK iOONT^PGRCE \CONT_RANK i * " ' 3 " l o f l ] " 6 " . ' 6 6 f 6 ; a o d ] o \ o".'s7| i"toj"" c s i ; a-ist '""o'.'doy-~oMr aooT" b77i'!"' . o.ooi ' a'06!" oVisi 6V17T " "aooT ZMI'" ""'0"29j " ""32:64?" 4 X 3 1 ! "0>2: Si" i i i "b"iaT '""'61 7" 'b.'59l i! 22 4 61 s "0.42I ii 83 "047]' ' " ' 1! 35 "bis'iT ZZZZQZZ'Z 7 "b!42! ' 1; 35' 6.68; 2i 65 "6.73! '" ""2! 14* b.&si 2i "" is 0.60* i; " ~ 4 3 "aiiit bt 6 "0.83] 2; 29 O.liii 6i 4' 077; 7.1 27~ "o'Mii\" 2: ""31 "awt 2: 122 "007? bi 6 "a'72j 2! 53 "'044! 1; 0 ' 0.83] 22" 072] 'ZZZZZMZ"" 6 "b"43i 46 *4 6j 42 Final Complete IMSf .ittftiS^.. _ .PflOV.O)OE:061.CDO!; COU.Ca£;CASCS.10M C&KJVm XFWJF&t [UflJM. [ISfJVJ* OJH.MW :Z.MWC»:OOW.RW ;WlWJiCW:«-J^ .M^VttAn;^ nnin'cw' Fig 4.9 Diagram of Data Unions Creating Final Data Table 75 Data Analysis This commentary on the methodology now continues by describing how hotspot zones were created, and also how the spatial correlation analysis was carried out. 1) Hotspots At this point, the methodology by itself has not produced anything 'original' in the sense that existing data had been merely re-classified by different spatial aggregates. Although these data might not have been mapped, experts working with the data sets know intuitively which regions are most afflicted by the particular threat that they study. However, due to disciplinary isolation, lack of inter-agency collaboration, and difficulty in translating data sets, this information rarely gets compiled and almost never directly compared. Knowing which regions of a country are subject to which and how many human security threats would be invaluable insight for policy experts. As explained earlier, I call these regions of aggregated human insecurity 'hotspots'. Process In order to create a hotspot map, we must determine how many 'level two' threats are present in any given commune of Cambodia. This is done using the COUNT IF function in ArcView. By asking the database to count all records in a field that are equal to 2, we can determine how many threat level 2s (Highs) are in each commune, (see Table 4.5) 76 Table 4.5 Example of 'Hotspot' Table COM .C00E M A X L D E h M A X J d A l M A X _ G U r M A K D O M V MA>JNF_RA M A X _ H f t V _ R A M A X _ " f B _ R A N M A X . F L O M A K P f i C M A X _ P O \ M A X _ C O r M A X _ P L G M A X - A I D S _ R 1'nH.rnhE.A 99 0 0 0 0 10201 2 2 1 2 2 0 0 0 1 2 10202 2 2 1 2 2 b 0 1 1 2 10203 2 2 1 2 2 b 0 1 1 2 10204 2 2 1 2 2 0 0 1 1 2 10205 2 2 1 2 2 b 0 0 1 2 10206 2 2 1 2 2 b 0 0 1 2 10207 2 2 1 2 2 b 0 0 0 2 10208 2 2 1 2 2 b 0 0 1 2 5 10209 2 2 t 2 2 0 0 0 1 2 /< 5 10210 2 2 1 2 2 0 0 ' 1 0 2 10211 2 2 1 2 2 b 0 1 1 2 10212 2 2 1 2 2 0 0 0 1 2 10213 2 2 1 2 2 b 0 0 1 2 10301 2 2 1 2 0 b 0 1 1 2 4 10302 2 2 1 2 0 b 0 1 1 2 10303 2 2 1 2 b b 0 1 0 2 10304 2 2 1 2 0 0 0 0 1 2 10305 2 2 1 2 Q b 0 0 1 2 % 4 The new number given to each commune in the right hand column of Table 4.5 (number of high threats) can now be mapped in the GIS. As well as being able to see clearly which regions are less secure/highly vulnerable (see Fig 4.10 on next page), any of the data used to come to this final assessment are easily available in tabular format. We can determine, what threats resulted in the presence of a hotspot and then we can analyze the original (pre-ranking) data sets. Fig. 4.10 Hotspot Map (next page) 11 2) Spatial Correlations As noted earlier, the methodology used also facilitates the study of correlations between threat variables through a spatial analysis. Just as all variables were overlaid together in the hotspot analysis, any combination of threat can be analyzed together to look for spatial relationships. For example, spatial correlation analysis can be used to try and answer the following two hypothetical questions (of course these are examples from a possible 13x13 matrix of possible correlations): 1. Are areas of high landmine contamination subject to an unusually high prevalence of other human security threats, i.e. poverty, disease, violence? 2. Under what conditions, (i.e. high Dengue Fever), is it likely to find a high poverty threat? 1. Landmine Area Analysis Landmines were chosen for a spatial correlation analysis as they remain a devastating problem in Cambodia. There remain almost 3000 landmine contaminated fields and approximately 1000 people are killed a year from either landmines or UXO's (CMVIS Landmine Monitor Victims Report, 2002). Although substantial landmine removal programs are in place, only several dozen fields are de-mined per year. The problem will certainly be acute for decades to come. Although landmines are a devastating human security threat to some Cambodians, the problem is highly spatially determined - for instance, someone living in urban Phnom Penh has little fear of stepping on an unexploded ordinance. For this reason, spatial analysis could be very useful in demonstrating corresponding threats that are present in rural regions of high landmine contamination. For this analysis, I tested the percentage of landmine contaminated villages at the district level, with high threats from the other 12 indicators used in the hotspot analysis59. By using a series of logic questions (If A 5 9 There has been significant deliberation as to what is the best way of capturing the human threat of landmines. One option is to calculate the total area of contaminated land and use a percentage of a political units contaminated area. This, however, is highly skewed by dramatically varying district and commune 79 and B , then C) a series of new maps were created showing the degree of spatial correlation between landmines and each the other human security threats. This is done by querying the final data table for districts with high landmine threat, A N D high poverty threat, high malaria threat and so on. Table 4.6 shows the compiled results of 12 separate queries (i.e. high landmine contamination and high threat x). The first column shows the threat x that is being tested with landmine contaminated land. In brackets are the total number of districts with high rankings for each threat. The second column is the number of districts that "passed" the If A and B query - i.e. districts with both high land mine contamination and high threat (x). The third column is the number of correlating districts as a percent of all high contaminated districts. The final column is the national probability of that threat being high in any given district. What is important for this correlation analysis is having both a high percentage of correlation between landmine threat and one of the other 12 variables, as well as a significantly greater correlation than the national probability for each threat, i.e. A high difference between column 3 and column 4. For example, if 100% of all high contaminadon threat districts also had high poverty, but, all districts in the country had high poverty, then the correlation would be null. Table 4.6 Correlation Table 1 Number of correlating districts % of Total Cont Natn'l Probability (total high/186) Cont (37) PLG (72) 8 22 39 TB (75) 8 22 40 AIDS (53) 13 35 Dengue (50) 22 59 27 Malaria (39) 10 27 Gun (52) 15 41 DomV (30) 12 Infant Mortality (113) 14 38 61 HRV (72) 15 41 39 Flood (15) 3 8 8 Drought (11) 2 5 6 sizes and populations. By using village contamination, instead of area, there is a necessary link between people (particularly farmers) and the landmines. 80 Results A n often stated hypothesis is that landmines cause poverty. In this case, the analysis showed that 23, or 62%, of the 37 heavily landmined districts also had a high poverty threat. Although this is a very high correlation, the national probability is 55%. There is certainly a relationship, but in a country with as much serious poverty as Cambodia, there many other significant causes (see poverty analysis below). What is particularly interesting is the high correlation between disease, (specifically HIV/AIDS and Dengue Fever), and violence, (injuries by guns and domestic violence), with high landmine threat. For example, Table 4.6 indicates that 59% of contaminated districts also have high dengue threat, where as the national probability is 27%. Also, 41% of contaminated districts have a high gun threat and 32% have a high rate of domestic violence, compared with the national probability of 28% and 16% respectively. These figures, while not absolutely conclusive, do show a higher than average probability of poverty, disease and violence being associated with the heavily landmined areas of Cambodia. 2. Poverty Area Analysis Poverty affects a vast number of Cambodians. The country is so poor that the poverty line is calculated as a monetary equivalence of calorie intake rather than individual or household income. As so few people even earn an income, their actual finances are irrelevant. Impoverished people in Cambodia either eat what they grow, use all of their money on food, or scavenge60. There are three ways of analyzing the link between poverty and the human security variables used in this study using the GIS database. First, by looking at the probability of high poverty rates in communes of high threat (x), second by doing the inverse and looking at the probability of a high threat (x) in communes of high poverty, and third by doing a visual analysis of map overlays. The poverty line is approx. 1300 reil (approx $1.50 CAN) per person per day on monetary equivalent of calorie intake (interview notes). 81 First, by comparing incidence of, or probability of, poverty in each threat category as compared with the national high poverty threat probability, an indication can be given as to what threats, if any, might exacerbate, lead to or cause poverty61. Again, we are looking for significant differences between the probability of poverty in regions of high threat (x) and the national average. The table below shows the results from 12 queries using the same method as in the landmine analysis. The first column represents the number of communes with both high poverty and threat (x). The second column is the percentage of high poverty communes in all communes with high threat (x). The final is the national average of high poverty threat occurrence at the commune level. Table 4.7 Correlation Table 2 Threat (x) Number of Correlating Communes Prob. of poverty within threat (x) communes Prob. of Poverty Nationally Cont (221) 90 PLG (89) 26 29 28 TB (762) 173 23 28 AIDS (440) 101 23 28 Dengue (350) 137 2f Malaria (209) 28 13 Gun (452) 82 18 28 DomV (208) 110 28 Infant Mortality (1030) 272 26 28 HRV (695) 183 26 28 Flood (149) 29 19 28 Drought (123) 32 26 28 The five rows shaded In Table 4.7 stand out as showing relationships that differ significantly from the national average. First, the probability of a commune having a high poverty threat is greater in communes of high landmine contamination62, dengue fever and domestic violence. Second, malaria and gun injuries are lower in regions of high poverty than the national average. For gun injuries this could be explainable by 6 1 Determining which of these is taking place is out side the scope of this project. I believe, however, that were this type of analysis to be done in a multi-country case study, these types of question could be successfully examined. 6 2 As in the previous analysis there is a strong correlation between landmines and poverty. The slightly higher rate of correlation in this analysis is probably due to the fact that it is being done with commune rather than district level data. For the landmine contamination analysis district level data was most accurate for representing the independent variable, so all data, including poverty, was aggregated to this level. The poverty data, however, is most accurate at the commune level, so was kept at this resolution for the poverty analysis. 82 its higher correlation with landmine contamination rather than poverty per se. The low association of high malaria threat with poverty, however, is perhaps surprising and indicates that malaria must be primarily located in regions with neither high landmine contamination nor poverty. It is also interesting that the poverty threat is not significantly high in regions with high flood affected communities. Another scenario addressing this will be looked at in the third poverty analysis. A second way of analyzing poverty and human security indicators in this data set is summarized in Table 4.8. Rather than looking at high poverty in threat (x) this analysis does the inverse, looking at the presence of threat (x) in regions of high poverty. Accordingly, this will indicate the inverse of Table 4.7, i.e. whether the presence of threat (x), leads to, exacerbates or causes poverty. Table 4.8 Correlation Table 3 Threat (x) Number of Correlating Communes Prob. of threat (x) within Poverty Communes (451) Prob. Of Threat Nationally Cont (221) 90 PLG (89) 26 6 5 TB(762) 173 38 47 AIDS (440) 101 22 27 Dengue (350) 137 21 Malaria (209) 28 6 13 Gun (452) 82 18 28 DomV (208) 110 'j j.„<r<^ , 13 Infant Mortality (1030) 272 60 63 HRV (695) 183 41 43 Flood (149) 29 6 9 Drought (123) 32 7 8 Again, the same three threats (shaded in Table 4.8) show significance in this analysis. Contaminated land, dengue fever and domestic violence all have a higher correlation with poverty than the national probability. A third way of spatially analyzing the poverty data is through a visual analysis of map overlays. As the flooding data (flood affected communities) has thus far shown no abnormal correlation with poverty, the relationship can be analyzed using a new data set - water levels of the 2000 Cambodia flood. 83 The following three maps ( Figs 4.11 to 4.13) depict the overlay process from the original radarsat image of the flood to the final map showing areas of high, medium and low poverty threat overlaid with the peak water levels of the 2000 flood. 84 Fig 4.1 1 Radarsat Flood Map Source for Radarsat Image: Mekong River Commission € 4 This final overlay map clearly shows a spatial relationship between poverty and flooding, especially along the Tonle Sap basin. The severe poverty, present in almost all locations of flooding, extends beyond the flooding in a consistent and seemingly causal manner. This could be due to adjacent and dependant food supplies being affected in flood regions. This shows how powerful simply looking at data can be. Base Map Overlays Another unique ability of Geographic Information Systems is their ability to qualify analysis by adding some of parameters contained in base maps. This data could be any topographic feature, such as roads, villages, cities, rivers, borders, or images of detailed maps. Or, it could be any information relevant to the analysis, such as US bombing locations, landmine contaminated areas, the location of Khmer Rouge hideouts, or the areas subjected to high levels of violence during the war6 3. Adding this type of information creates a powerful tool in facilitating development assistance and humanitarian relief efforts. This is because regions of insecurity can be isolated and logistically relevant features such as roads, population rivers etc, can be visualized. The following maps (Figs 4.17 to 4.23) give examples of base maps that could be overlaid with any of the threat maps, including the final hotspot analysis. The last two data sets are collected by the Cambodian Genocide Program at Yale. There may be opportunity for future collaboration. 87 Fig 4.14 Example of Road Map Fig 4.18 Example of Topographic Map Overlay (above) F ig 4.1 9 Example of Topographic Map Overlay Zoom 1 (below) C a m b o d i a Topograph ic B a s e M a p s (Zoom ?) • Index liriHgo Map Sheet 359342 .'/ap stwels SO.stip 90 Cambodia Topographic Base Maps (Zoom 2) Index Image Map Sriwi *59342 Map sr » « 3 SO sup Fig. 4.20 Example o f Topographic Map Overlay Zoom 2 (above) Fig. 4.21 Example o f Topographic Map Overlay Zoom 3 (below) Cambodia Topographic Base Maps (Zoom 3) 91 4.4.4 Conclusions of Analysis The methodology described in Chapter 3 has proved to be both feasible and informative. A l l stages of the analysis revealed aspects of Cambodian vulnerability that are innovative and relevant. First, although Cambodia is not at war, either with another country or within its borders, it is not secure. This is clear by its very low ranking on all global development indexes. This insecurity is caused by the presence of 17 broadly defined human security threats grouped into 6 human security categories. These threats, identified by a series of interviews with disciplinary experts, are specific to Cambodia. Another country, with similar levels of insecurity, could be subject to a completely different threat list. Data detailing most of these threats is available and relatively reliable. Along with necessary geographic base maps, data was collected, organized in a database and spatially referenced. Second, a Geographic Information System was used to map and spatially analyze the threat data. A hotspot map clearly reveals areas of high aggregated vulnerability. These regions, subject to multiple high (threat level 2) human security threats, warrant immediate attention and further analysis. Third, correlation analysis revealed a significant relationship between high landmine contamination and high poverty, dengue fever, HIV/AIDS, domestic violence and gun injuries. In addition, poverty was spatially correlated with dengue fever, domestic violence and landmines. While these correlations do not imply necessary causality, they do show a degree of significance that warrants further inquiry. This spatial significance is also apparent using visual rather than mathematical overlays. In the GIS, base maps can be added to any layer of the analysis to facilitate development and humanitarian relief efforts. The final chapter of this thesis will return briefly to the relationship between geography and human security, and look at the role of this thesis and the methodology it prescribes, to the theory and practice of measuring human security. 94 Chapter 5 Ways Forward 5.1 Human Security and Geography There is a clear mutually beneficial relationship between the concept of human security and the field of human geography. Human security, as considered in this thesis, represents an overarching framework in which any issue, as long as it represents a clear threat to the vital core of the individual, may be included. The consequence of this inclusiveness is that analytic inquiry becomes very difficult, both do to the shear number of possible threats and the inherent interdisciplinarity required to compare them. The solutions to these dilemmas were covered at length in this thesis. Using space as a common denominator and regional relevance as a key indicator selection criteria, were among those prescribed. These were, however, only apparent, once a distinctly geographic perspective was used. Seeing space as an attribute of every security threat, whether it be a war, a disease or a flood, provided clear answers to the difficulties prior researchers have faced. It also became clear that the broad, interdisciplinary nature of Geography, allowing collaborative research between the natural and social sciences, is ideally suited to house human security research. This being said, this symbiotic aspect of geography is underused and undervalued. There is surprisingly little cross-over research conducted in geography departments65, funding driven research being increasingly forced to specialize. For this reason, an interdisciplinary conceptual framework such as human security offers a unique opportunity for proponents of broad geographic research. 5.2 Implications of the Cambodia Case Study for Human Security 5.2.1 Theoretical Implications Proponents of human security need to swiftly and directly address their critics. Prominent critiques are in general neither well formulated, nor impervious to fairly straightforward retort. Most revolve around a defense of the status quo, of an international order that has recently seen premeditated dissolution from the 6 5 A notable exception being the study of the mutual vulnerability between the environment and humans. 95 very supporters of the traditional security paradigm. A defense of human security complements a defense for international cooperation and humanitarian assistance and can be promoted in tandem with other broadening security concepts. Human security is both acutely prescient and timely: proponents must begin to defend it as such. Coinciding with a strong conceptual defense, proponents of human security must continue to push human security initiatives forward. The International Conventional for the Ban of Landmines (ICBL) and the International Criminal Court (ICC) were a start, but there are many other prominent international concerns that fall under the possible rubric of a human security agenda. In order to properly and responsibly address the real threats people face, a broad conceptualization of human security, one that includes all serious threats to the individual, must be used. Isolating and propagating these conceivably vast number of issues should be aided by measuring methodologies. In this thesis I have argued that measuring human security, particularly at the intra-state level, has the potential to highlight key threats, bringing the issues that truly affect peoples' lives to the fore of international policy. The methodology must be both broad, to capture all possible threats, and accurate, to meet rigorous analytical critique. 5.2.2 Implications For Measuring Human Security A methodology was put forward in this paper that is both conceptually broad and analytically rigorous. Based on the foundation of a geographic information system, it has the ability to organize and analyze vast amounts of interdisciplinary data. Its feasibility, however, was initially questioned. Critics alluded to issues of data availability and quality as well as to possible local reluctance to data sharing and collaboration. Although certainly legitimate concerns, a highly successful field trial in Cambodia showed them to be surmountable. 96 Based on the Cambodian case study, several conclusions as to the methodology's appeal, procedure and future applications can be drawn. On the Appeal of Human Security First and foremost, there was broad based enthusiasm in Cambodia, amongst researchers, NGO's and Government Ministries, for the concept of human security, the utility of measuring it, and the possible applications of GIS. As a concept, human security is relatively new to the Cambodian research and N G O community. Although regionally, there has been much discussion of human security at both a government and research level, in Thailand and Japan for example, it has not yet reached mainstream discourse in the less developed areas of Asia. This being said, when confronted with the principles of the concept, protecting the individuals from non-traditional harms, Cambodians and ex-pats working in Cambodia found a particular resonance. This is due in part to the fact that since the end of the war, few people in Cambodia have died in violent conflicts but fewer still would consider themselves secure. The idea of measuring human security initially raises skepticism in the Cambodian research community. Measuring implies a degree of empirical certainty that makes many, particularly those outside of the natural and positivistic social sciences, uneasy. Once the methodology was outlined, however, most of the people I interviewed saw its utility (see more on possible applications below). Geographic Information Systems are also garnering significant attention in Cambodia. Their utility is increasingly being recognized and many organizations are building a technical capacity to use them. There is a network emerging of GIS practitioners and many of the map formats are being standardized - an essential prerequisite for effective broad based application. 97 O n the Procedure of Measuring Human Security Although the overall outcome of the methodological trial was successful, much was learned in the process. Conclusions regarding the methodological process and data issues will be valuable to future studies. Methodology First, a methodology ideally requiring the large number of interviews, panel discussion, and data collection as carried out in this study, should be done by more than one researcher. Although it proved to be possible, ideally it would be conducted by a team of researchers, including a GIS technician. The team could be modeled after the FEWER process (Forum for Early Warning and Emergency Response)66. Second, the threat assessment and data collection stages of the methodology should be autonomous. Due to time and resource constraints, they were conducted simultaneously in this study, but this does have the potential to bias threat determination and data selection. Those interviewed in the threat assessment stage could stress threats for which they have collected data. Third, final threat rankings (the nominal scale of 0, 1, 2), as discussed in Chapter 4 could be determined by disciplinary experts in or outside of the case study country. Although I believe that the method used in this case study remains valid, there is another option. The threat levels for each indicator could be determined based on international standards chosen by experts in each threat. For example, an epidemiologist could determine at what point malaria cases per 100,000 is a low and high threat. Data and Scale The primary concern throughout the process of developing and testing the methodology has regarded data availability and quality. Regarding data availability, the methodology shows that by only looking at regionally relevant threats, the potentially infinite list of possible variables can be reduced to a manageable This group does country assessments by first collecting data and contacts from London country, they then do a series of field studies to the study country, in which they hold expert workshops, interviews and collect data. i 98 number. The seemingly unmanageable "broad conception" of human security rendered analytically conceivable67. As for these dozen or so regionally relevant threats, sufficient data are likely available for all but the poorest and most war-torn countries. There may not be data for all conceivable human security threats but data are likely available for those affecting a specific county. This being said, caution is necessary regarding data quality. What is critical is accepting a degree of subjectivity in the data selection. The researchers must make an expert advised, qualitative decision as to the reliability of each data set. Although potentially biased or inaccurate, the wealth of locally procured and disaggregated data must be used by the broader research and development community. For this to occur, a small degree of error should be accepted. Such data, including a marginal degree of error, are likely much more accurate than the generalized national level alternatives, particularly for isolating specific areas of high vulnerability. Because of its lack of disagregation, a national poverty rate tells us little as to who is suffering, as a national flooding statistics reveals nothing as to where it has occurred and who has been harmed. On Future Applications of the Methodology The possible future applications for this methodology are vast, for Cambodia, for the international development and humanitarian relief community and for academia. Most importantly, the data and the completed GIS tool will be returned to Cambodia to be used by NGO's, IOs and Government Ministries. Much of this data has never been mapped and almost none of it has been spatially analyzed. The results will be useful to many68. Again, if a methodology calls for data on 200 threats in every country, they will be forced to use very vague, and some would argue meaningless, national level statistics. 68 For example, although the WHO collects data on malaria, TB, dengue fever and HIV/AIDS, they have never been mapped together, an exercise that will be valuable to the organization. 99 Ideally, the next stage in the Cambodian human security assessment would consist of a qualitative "ground-truthing" of the data. Hotspot regions could be assessed to determine whether they truly do suffer from a relevant difference in vulnerability. Regions that have shown significant relationships in the correlation analysis would also be evaluated by experts in the field On a much broader scale, the methodology has two paralleling potential applications, both requiring multi-country case studies. First, as a logistical tool for development planning and humanitarian assistance, a human security GIS would be invaluable. Knowing what threats are present in a country and exactly where they are prevalent would help in both the conception and execution of programs and aid. The methodology also facilitates the type of data and information sharing that rarely occurs in this policy community. For agencies in a region to have access to each others data and knowledge of all development initiatives would provide a level of streamlining that is desperately needed. Humanitarian assistance should not be a competitive isolationist enterprise, it must cease to be operated as such and work toward collaboration as opposed to antagonism. CIDA, for example, could commission a human security assessment of a country for which they were planning significant bi-lateral assistance. This assessment, along with the GIS database would help them both target the most vulnerable peoples, choose the most effective programming and avoid many of the inevitable pitfalls of development planning, such as project redundancy and lack of inter-agency collaboration. At a time of immediate post-conflict assistance this information would be ideal. As aid floods into a country there is often little thought as to exactly what is needed and where. More than often the people who need help most, the most insecure, never receive the assistance they desperately require69. The second future application of this methodology is academic. The study of the causes and consequences of war is a burgeoning inter-disciplinary endeavor. Spatial analysis should be part of this dialogue. 6 9 As a testament to this, one only need look at post conflict Afghanistan. NGOs clamoring for media coverage failed to address the majority of the people truly suffering. Horror stories of thousands of security blankets arriving at a village desperately needing food and water, are not uncommon. 100 The causes of war, particularly intra-state conflict, have undergone dramatic re-evaluation. The triggers of complex regional instability and violence often involve an intersection of policy, identity, resource location, poverty, disasters and vulnerability. These relationships, however, have generally only been studied at the national level and in disciplinary isolation - for instance economists studying poverty and conflict, geographers studying resources and conflict, political scientist studying governance and conflict and so on. An intra-state spatial analysis of all of these potential conflict-causing variables could reveal correlation and perhaps causation that has thus far gone undetected. This type of analysis could also be conducted as a historical exercise, i.e. under what conditions did regional conflict break out7 0? The long term socio-economic consequences of war also warrant further investigation and again, spatial analysis of human security indicators provides a unique facilitating mechanism. In an era of high altitude bombing and heavy landmine use, the past 20 years has resulted in devastation, long after the conflicts have ended. The long term effects of landmines and unexploded ordinances (UXOs) can be crippling to a country already ravaged by the immediate impacts of war. Showing exactly what and how serious these socio-economic impacts are, however, would be invaluable to the promotion of a human security agenda. A multi-country GIS analysis could be developed from this project's brief examination of the spatial correlation of landmines and human insecurities. Does heavy landmine contamination always result in higher incidences of poverty, dengue fever and violence, 20 years after their deployment (as is shown in this study)? Do the UXOs from cluster and carpet bombing leave a long lasting legacy of insecurity, as they do in Cambodia? These types of questions warrant analysis. I hope that the development and preliminary testing of this methodology has shed light on the relevance of human security, the necessity of a broad definition, the utility in measuring it and on the value of spatial analysis. For it to be used to assist a country as troubled as Cambodia is sufficient, for it to see broader application in the development and academic community is a long term goal. 7 0 For a brief discussion of the causes of war literature, see Appendix 4 101 Works Cited Acharya, Amitav. "Human Security: What Kind for the Asia Pacific? What Options?". Presented at the 15 t h Asia Pacific Roundtable, Confidence Building and Conflict Reduction. June 2001, Kuala Lumpur. Antenucci, J., Brown, K. , Croswell, P., Kevany, M . Geographic Information Systems. Van Nostrand Reinhold, 1991. Axworthy, Lloyd. "Human Security and Global Governance: Putting People First". Global Governance 22:1 (2001): 19 Bain, William. "The Tyranny of Benevolence: National Security, Human Security, and the Practice of Statecraft". Global Society 15:2 (2001): 176-294. Bajpai, Kanti. "Human Security: Concept and Measurement" Kroc Institute Occasional Paper #19. 2000. Barrows, H . 1923. Geography as Human Ecology. Annals of the Association of Human Geographers 13: 1-14 Baylis, John. "International Security in the Post-Cold War Era." In Baylis, John and Steve Smith (eds.): The Globalization of World Polititics. Oxford: Oxford University Press, 1997 Bernhardsen, T. Geographic Information Systems an Introduction, 2 n d Edition. John Wiley and Sons Inc., 1999. Blaikie, P., and Brookfield, H . 1987. Land Degradation and Society. London: Methuen. Brauer, Jurgen. "War and Nature: The Problem of Data and Data Collection" Paper for the Uppsalla Conference on Conflict Data, June, 2001. http://www.pcr.uu.se/ident.html (Downloaded 2002) Chandler, David. A History of Cambodia. Boulder, CO: Westview Press, 2000. Collier, Paul A. Hoeffler. "Data Issues in the Study of Conflict." Paper for the Uppsalla Conference on Conflict Data, June, 2001. http://www.pcr.uu.se/ident.html (Downloaded 2002) Collier, Paul . "Doing Well out of War," Conference on Economic Agendas in Civi l Wars. London. April 26-27, 1999. www.worldbank.org/research/conflict/papers/econagendas.pdf (Downloaded (2001) Collier, Paul. "Economic Causes of Civi l Conflict and Their Implications for Policy". 2000. www.worldbank.org/research/conflict/papers/civilconvlict.pdf (Downloaded 2001) Covello, V . , Mumpower, J. Risk Analysis and Risk Managemnt: An Historical Perspective. In Cutter, S.L. ed Environment, Risks and Hazards. Pretice-Hall: New Jersey 1994. Cutter, S.L., and Solecki, W.D. 1989. The Natural Pattern of Airborne Toxic Releases. The Professional Geographer 41: 149-61 Cutter, S.L.; Mitchell, J.T.; Scott, M.S. 2000. Revealing the Vulnerability of People and Places: A Case Study of Georgetown County, South Carolina. Annals of the Association of the American Geographer 90: 713-37. "Definitions of Human Security." Harvard University Program on Humanitarian Policy and Conflict Research Website, http://www.hsph.harvard.edu/hpcr/events/hsworkshop/comparison_definitions.pdf Denis, Michel (1991): Image and Cognition. New York: Harvester Wheatsheaf. 102 Denis, Michel et al (2001): Imagery, Language and Visuo-Spatial Thinking. Hove: Psychology Press Ltd. Deudney, Daniel. 'The Case Against Linking Environmental Degradations and National Security". Millennium 19 (1990): 464. Foong Khong, Yuen. "Human Security: A Shotgun Approach to Alleviating Human Misery?". Global Governance 7:3 (2001): 231. Gurr, Ted Robert; Marshall, Monty; and Khosla, Deepa. "Peace and Conflict: A Global Survey" 2001. pp. 1-21. www.bsos.umd.edu/cidcm.CIDCMpeace.pdf (Downloaded 2002) Haftendorn, H . 'The Security Puzzle: Theory Building and Discipline Building in International Security". International Security Quarterly. 35:1 (1991), 3-17. Hampson, Osier, Hay, J., Human Security: A Review of Scholarly Literature. Paper presented to the Canadian Consortium on Human Security Annual Meeting, Ottawa, April 2002. "Harvard Program on Humanitarian Policy and Conflict Research: Human Security Report." (2001). Harvard Program on Humanitarian Policy and Conflict Research. Get this website Heywood, I., Cornelius, S., Carver, S. An Introduction to Geographical Information Systems. Addison Wesley Longman Limited, 1998 Hobbes, Thomas. The Leviathan. Penguin Books, reprinted 1985. Homer-Dixon, Thomas. "Environmental Scarcities and Violent Conflict: Evidence From Cases", International Security, 19:1 (1994): 5-40. Kasperson, R.E.; Renn, O.; Slovic, P.; Brown, H.S.; Emel, J.; Goble, R,; Kasperson, J.X.; and Ratick, S. 1988. The social Amplification of Risk: A Conceptual Framework. Risk Analysis 8: 177-87. Kates, R., and Burton, I. 1986. Geography, Resources and Environment, vol. 1. Chicago: University Press. Keohane, Robert. After Hegemony: Cooperation and Discord in the World Political Economy. Princeton: Princeton University Press, 1984 Keohane, Robert and Lisa Martin. "The Promise of Institutionalist Theory". International Security 20:1 (1995). K i m , Woosang. "Human Security Concerns in Global Politics". Prepared for the 9 t h Council for Security Cooperation in the Asia Pacific Working Group Meeting on Comprehensive and Cooperative Security, Novotel Wellington, New Zealand, April, 2001. King, G., Murray, C. "Rethinking Human Security." Harvard University Program on Humanitarian Policy and Conflict Research. (2000) Klugman, J. "Social and Economic Policies to Prevent Complex Humanitarian Emergencies". Policy Brief No. 2, UNU/WIDER, pp. 1-5 Krause, Keith. "Une Approch Critique de la Securite Humaine". Program D'etude Strategique et se Securite International, Institute Universale de Haute Etude International, Geneve. Septembre 2000. Leaning, Jennifer., Arie, S. "Human Security: A Framework for Assessment In Conflict and Transition". Prepared for USAID (Africa Bureau)/Tulane CERTI. (2000). To be published in the Working Paper Series, Harvard Center for Population and Development Studies. 103 Lonergan, Steve, Gustavson, K., Carter, B . 'The Development of an Index of Human Insecurity". Global Environmental Change and Human Security Project. International Human Dimensions Programme On Global Environmental Change. Research Report. Draft. September 1, 2000 Mack, Andrew. (2002). "Data Issues." Forthcoming. MacFarelane, S. " A Critical History of the U N and "Human Security". www.liucentre.ubc.ca/huamnsecurity/course/MacFarlane.htm. (Downloaded 2001) Mearsheimer, John J. "The False Promise of International Institutions." International Security 19:3 (1994/95) Miall , Hugh. "Data Requirements for Conflict Prevention." Paper for the Uppsalla Conference on Conflict Data, June, 2001. http://www.pcr.uu.se/ident.html (Downloaded 2002) Mitchell, J.T.; Scott, M.S.; Thomas, D.S.K.; Cutler, M . ; Putman, P.D.; Collins, R.F.; and Cutter, S.L. 1997. Mitigating Against Disaster: Assessing Hazard Vulnerability at the Local Level. GIS/LIS Proceedings 563-71. Bethesda, M D : A C S M , ASPRS, A M / F M , AAG,URISA. Nef, Jorge. Human Security and Mutual Vulnerability. The Global Political Economy of Development and Underdevelopment. Second Edition. Ottawa: International Development Research Center, 1999. "New Dimensions of Human Security." Human Development Report. United Nationns Developmet Program. New York: Oxford University press. Chapter 2 , p.22-25, 1994. Newman, Edward. "Human Security and Constructivism". International Studies Perspectives 2 (2001): 239-251. Owen, Taylor. 2002. "Body Count: Rationale and Methodologies for Measuring Human Security". Human Security Bulletin. October, Vol.1 No. 3. Commissioned by the Canadian Consortium on Human Security and used at the United Nations University and World Peace Forum workshop "Culture of Peace Indicators." U N U Tokyo, Sept. 2002. Palm, R. 1990. Natural Hazards: An Integrated Framework for Research and Planning. Baltimore; John Hopkins University Press. Paris, Roland. "Human Security: Paradigm Shift or Hot Air?". International Security 26:2 (2001): 87-102. Rothschild, Emma: "What is Security?" Daedalus 124:43 (1995): 53-90. Richardson, John T.E. (1999): Imagery. Hove: Psychology Press Ltd. Rummel, Rudolf. "Testing Whether Freedom Predicts Human Security and Violence". http://www. mega. nu:8080/ampp/rummel/wf.appendix. htm Spence, Robert (2001): Information Visualization. Harlow: Pearson Education Limited. Starr, Harvey., Bain, W., "The Application of Geographic Informations Systems (GIS) to International Studies". International Studies Notes 20:2 (1995): 1-8. "The Index of Human Insecurity." AVISO Bulletin. Issue No. 6. (2000) www.ghechs.org/Aviso/ (Downloaded 2001) "The Responsibility to Protect". Report from the International Commission on Intervention and State Sovereignty. Available at www.iciss.gc.ca/report-e.asp (Downloaded 2002) 104 Thomas, Caroline, "Global Governance, Development and Human Security: Exploring the Links". Third World Quarterly 22:1 (2001): 159-175. Ullman, R. "Redefining Security". International Security 8:1 (1983): 129-153. Waltz, Kenneth. Theory of International Politics. New York: Random House, 1979. Wendt, Alexander. "Anarchy is What States Make of It: The Social Construction of Power Politics." International Organization 46:2 (1992). White, G.F. 1945. Human Adjustments to Floods. Department of Geography Research Paper 29. Chicago: University Press. White, G.F. 1964. Choice Adjustments to Floods. Department of Geography Research Paper 93. Chicago: University Press. Wilhite, D., and Easterling, W. eds. 1987. Planning for Drought: Towards a Reduction of Social Vulnerability. Boulder, CO: Westview Press. 105 _o o -o o •g u s o D . O OH c =3 X \0*> led Appendix 2 Dear CDRI Colleagues, As you may already know, I am visiting CDRI as an intern from the University of British Columbia in Canada. My field of study is measuring and mapping threats to human security. I am interested in the possibility of identifying, locating and overlaying the most serious of threats facing Cambodians. Security studies have traditionally focused on the integrity of the state - with threats being interstate war, nuclear proliferation and shifts in power balances. Human Security, by focusing on the security of the individual rather than the state, broadens this list of threats to include, the environment, the economy, health, food, personal safety and politics etc. Although there are many conceptions of human security, for my research I will be using the following definition: The objective of human security is to safeguard the vital core of all human lives from critical and pervasive threats. This definition limits human security to a focus on the most serious, very specific threats, those that take or seriously harm lives. This is important in order to differentiate the concept from such broad, catch all notions as human development, sustainable development and poverty alleviation. As much of my research will involve interviewing researchers, experts, NGO workers and Government officials in Phnom Penh and around Cambodia, I was hoping that you might be able to assist me in determining the most appropriate people to speak with and how best to get in touch with them. Ideally, I would like to speak with several people from each of the areas of potential human security threats. I will be asking them: 1) What they feel are the most immanent concerns in their category of security. 2) Are there certain issue that cause direct harm or kill individuals 3) What is the best way of representing these threats i.e., qualitative reports, quantitative data sets, maps etc 4) To what spatial scale is this data available, are there significant spatial variances in the level of threat throughout the country 5) Can they direct me to these data sources It would be of great assistance if you could kindly fill out the attached form. I have also included a column for Mapping or GIS projects that you may know of. Any advice or comments you have would be helpful. Thank you in advance for your guidance and I look forward to speaking with all of you on a much more personal basis. Sincerely, Taylor Owen 110 Comment Name of Organizations Name of T As Name of Officials Security Threat Economic e.g. very low income levels, very high unemployment Food e.g. low food production, food availability, or calorie intake Health e.g. proliferation of disease, poor access to health services, high maternal infant mortality Environmental e.g. Natural disasters, poor water quality, poor land use Personal e.g., the threat of violence (war, crime, rape, banditry, landmines etc.) Political e.g. torture, disappearance, human rights violations Mapping or Geographic Information System (GIS) Projects in Cambodia ro o o a c o c E o n +3 r-« c .2 o o DC «> X U l n o o a. so (0 c w Q. ZD O CD CO cz cc E ZD I l-t u o u. a. o H T3 C « C O S .a a a a < a a E o c o o UJ -o o o ti-ro a X ro c a E c o la c UJ w c (0 0> w c 4) E 3 U o Q w c> "> &_ « c o « « C o O r a | x UJ c w C M C a X .2 - J •=> > > o a. 't-is 2 | -§2 O §9 D) < a i c o •o O) 2. O 3 O u. a a TO c 0) a. S s m u E o c o u UJ TO 4) X c <u E c o > c UJ Appendix 4 Root Causes of Conflict We need to gain a better understanding as to the nature of the conflict. There has been much academic debate exploring the direct causal linkages between nontraditional security issues and the rise of conflict. Tad Homer-Dixon has done many case studies on the connections between environmental change or resource scarcity and intrastate conflict (Homer-Dixon: 1994). Gurr, Marshall and Khosla, have made the connection between strength and democracy and the propensity for conflict. They use the Polity ranking which grades countries from one to ten, 0 being autocracy, 6-10 being democracies (Polity web site). They find that throughout the 20th century democratic countries have had the least number of conflicts, autocracies have had significantly more and countries in transition have had the most (Gurr: 2001). Collier sees the rise of intrastate rebellion as being fueled by economic incentives. He sees rebels, not as freedom fighters, but as profiteers, similar to organized criminals. Based on strong economic statistics and theory, he has determined that greed is a much more reliable determinant for the rise of conflict than grievance. Grievance is often capitalized on by the rebels, but it is greed that initiates propensity for conflict (Collier: 1999 and 2000). Klugman, acknowledging the role of the environment, worsening economic conditions and crisis of state legitimacy, also adds horizontal inequality and external shock as the principal five causes of conflict (Klugman: 1999) Despite these significant case studies, however, no project has revealed a direct causal link. They have all established a relationship, but none have isolated one single variable as the determinant cause of conflict. The environment, the economy, the political structure, or the social conditions are clearly all factors in the rise of conflict but non is sufficient in and of itself to explain the phenomenon. An unfortunate implication of this is that it has been used my some to show weakness in the peace building agenda. It is my opinion that the literature has gone astray. That fact that no one condition will necessarily lead to conflict, does not rule out the contributing role of each and says nothing to the implications of several conditions being present in one location. It is the aggregated effect of human insecurities that I feel may be the best possible indicator for potential conflict. Poverty in and of itself may not necessarily lead to conflict, but combined with political repression and a recent environmental disaster, may significantly increase the regional propensity for violence. Even if the link with violence is not made explicit, knowing the specific sub national concerns of a country and the exact location of the threats would be invaluable to development assistance programs. It is one thing to know that a country has a high degree human insecurity, and there are many indexes to determine this, but knowing specific points of vulnerability or aggregated hardships 'hotspots' at a local level would provide the critical information much needed by both policy and development community. 113 Appendix 5 Data Tables Economic and Food Security Poverty- Percent of population below the poverty line (commune) G I S C O D E C O M M U N E % B e l o w P o v L ine 70302 T a k a e n 7 2 % 196365 T h m a K a e v I 1 8 % 130301 C h o a m K s a n t 59% 190301 P r e a e k M e a s 1 4 % 130302 Tuek K r a h a m 4 2 % 130304 R u m d a o h S r a e 4 7 % 130303 Pr ing T h u m 3 1 % 130207 K a m p o n g S r a l a u M u o y 4 2 % 220302 C h o n g K a l 6 8 % , 130203 S a n g k a e M u o y 7 3 % 130205 M l u P r e y M u o y 6 6 % 16766 K u m r u 6 0 % 136268 K a m p o n g Sra lau P i r 3 1 % 136262 C h h a e b P i r 6 3 % 160104 T a L a v 1 5 % 10704* T h m a P u o k 7 7 % 130402 K u l e a e n C h e u n g 6 3 % 171206 S l a e n g S p e a n I 8 4 % 160909 P h n u m K o k l_ 7 % 10705 K o u k K a k t h e n 7 2 % 130403 T h m e i ' 4 4 % 16763 P h u m T h m e i 8 3 % 16864 Svay C h e k 7 2 % 10807 Treas 4 3 % 190508 P r e a h R u m k e l 3 4 % 130703 C h h e a n M u k h 39% 130704 P o u 3 % 130705 P r a m e 2 6 % 160101 Mal ik 1 6 % . . .1628 Health Security Dengue Fever - new case per 100,000 (Province) GIS C O D E P R O V I N C E N e w C a s e s D e n g u e 1 B a n t e a y M e a n c h e y 268 22 O t d a r M e a n c h e y 3 3 3 17 S i e m r e a p 397 13 P r e a h V i h e a r 315 r 19 S t u e n g T r e n g 3 2 B a t t a m b a n g 25 24 Pai l in 43T 15 Pursat 162 9 K0T1 K o n g 41 5i K a m p o n g S p e u 26 7 i K a m p o t 14 2 3 ! K e p ~ " ~ " " " 1 2 7 " 21 T a k e o 39 2 0 ; S v a y R i e n g T . 8 Kanda l 56 14 Prey V e n g 8 16 Ratanak K in 323 11 Mondu l Kin ™ " . Z I Z ° 10 Kratie """" 16 3 K a m p o n g C h a m 54 1 2 i P h n o m P e n h 73 18:Sihanoukvi l le 19 6 i K a m p o n g Thorn 54 4: K a m p o n g C h h n a n g 47 Malaria - % of population at risk (Province) GIS CODE PROVINCE Z_POP_AT_R 1 Banteay Meanchey 13 22 Otdar Meanchey 34 17 Siemreap 15 13 Preah Vihear r 81 19 Stueng Treng 86 2 Battambang 5 24 Pailin r 951 15 Purs at 7 9 Koh Kong 68 5 Kampong Speu 10 7 " 2 3 Kampot 8 Kep 53 21 Takeo 0 20 Svay Rieng ol 8 kandai 0 14 Prey Veng 0 16 Ratanak Kiri 94 11 Mondul Kiri 78 10 Kratie 26 3 kampong Cham 24 12 Phnom Penh 0 r w Sihanouku'lle 40 6 Kampong Thorn 16 4 Kampong Chhnang 16 HIV/AIDS- % of infected domestic and commercial sex workers (Province) GIS CODE PROVINCE DCSW i 1 Banteay Meanchey 37.3% 1 2 Battambang 29.9% 3 Kampong Cham 29.3% 4 Kampong Chhnang 42.9% 5 J<ampong Speu 36.8% r ~ 6 Kampong Thorn i 28.6%: 7 Kampot 42.0%; 8 Kandai 25.6%) 23 Kep j 9 Koh Kong 53.6% I 10 kratie 32.1 %j 11 Mondul Kiri ; 22 Otdar Meanchey I ; 24 Pailin 37.9% 12 Phnom Penh 26.3% i 14 Prey Veng 17.4% 15 Pursat 58.6% 16 Ratanak Kiri " ' 21.7% 17 Siemreap 28.7% 18 Sihanoukvllle 22.0% 19 Stueng Treng 38.0%: 20 Svay Rieng 7.1%| 21 Takeo 11.5% : 13:Preah Vihear i 13.0%: TB- Cases per 100,000 (District) G I S C O D E DISTRICT TB_100000 109 Malai Oi 108 Svay Chek 188 107 Thma Puok T 9 T 106 Serei Saophoan 100 1 105 O u Chrov 36 104 Preah Netr Preah 98 103 Phnum Srok _ 164 " ~ " " ~ 81 102 Mongkol Borei 2205 Trapeang Prasat 1547 ~ 133 2204 Samraong 2203 Chong Kai 314 2202 1 Banteay Ampil , 6" 2201 Anlong Veaeng 156 1714, Varin 0 1713 Svay Leu Oj 1712, Srei Snam O 17111 Soutr Nikom 205 1710 Siem Reap 181 1709 Prasat Bakong O 1707 Puok 213 1706 Kralanh 377 1704 Chi Kraeng 112 1703 Banteay Srei 149 1 1702 Angkor Thum 0 1701 Angkor Chum 188 1307[Tbaeng Mean Choy 479 1306jSangkom Thmei i O 1305JRoviong ; 0 1304i Kuleaen i 0 . . . 1 8 5 ! Infant Mortality- (Province) GIS CODE PROVINCE INFANT_MORT_FV\TE 1 Banteay Meanchey 78.2 22 Otdar Meanchey 77.9 17 Siemreap 77.9 13 Preah Vihear 71.3 19 Stueng Treng ' 71.3 2 Battambang 98.0 24 Pailin 98.0 15 Pursat 139.4 9 Koh Kong 70.7 5 Kampong Speu 68.3 7 Kampot 100.4 23 Kep 100.4' 21 Takeo j 96.6' 20 Svay Rieng 102.0i 8 Kandai 89.2; 14 Prey Veng 111.0 16 Ratanak Kiri 169.8! 11 Mondul Kiri 169.8 To Kratie 3 Kampong Cham 107.9! 12 Phnom Penh 37.6! 18 Sihanoukvilie 100.4 6] Kampong Thorn I 64.5 41 Kampong Chhnang i 129.3! Personal Security Landmines - Percent of contaminated villages (district) Summarized up to the commune level from up from the village contamination file Was then used as a percent contaminate villages per district ^ C O D E COMMUNE NUM_OF_CONTVIL 70302 Takaen 12 190305 Tnma Kaev 3 130301 Choam Ksant 4 190301 Preaek Meas 4 130302 Tuek Kraham 6 130304 Rumdaoh Srae 4 130303 Pring Thum 2 130207 Kampong Sralau Muoy 5 . 220302 Chong Kal 7 130203 Sangkae Muoy 2 ^ 130205 Mlu Prey Muoy 3 10706 Kumru 7 130208 Kampong Sralau Pir 4 ' 130202 Chhaeb Pir 4 160104 Ta Lav 5 10704 Thma Puok "T 130402 Kuleaen Cheung 2 171206 Slaeng Spean 14 160909 Phnum Kok 5 10705 Kouk Kakthen 9 130403 thmei 5 10703 Phum Thmei 7 10804 Svay Chek 14 10807 Treas 9 190508 Preah Rumkel 8 130703 Chhean Mukh 3 130704 Pou "~A\ 130705 Prame 3 160101! Malik 4 ,,,1628 ; Small Arms-Percent of physical injuries due to a gun (Province) GIS CODE PROVINCE Gun Injuries 1 Banteay Meanchey 12.9 22 Otdar Meanchey 13.9 17 Siemreap 6.0 13 Preah Vihear 12.5 19 Stueng Treng 12.5 2 Battambang 5.3 24 Pailin 5.3 15 Pursat 8.8 9 Koh Kong 4.0 5 Kampong Speu 15.1 7 Kampot 9.4 23 Kep 9.4 21 Takeo 0.0 20 Svay Rieng 7.5 8 Kandal 7.1 14 Prey Veng 0.0 16 Ratanak Kiri 0.0 11 Mondul Kiri 0.0 Kratie 3 Kampong Cham 9.9 12 Phnom Penh 15.0 18 Sihanoukville 9.4 Kampong Thorn 9.3 4 Kampong Chhnang 23.3 Domestic Violence-Percent of Married Women Experiencing Domestic Violence (Province) jGIS CODE I PROVINCE % of Married Women Violence! IjBanteay Meanchey 225 Otdar Meanchey 17J Siemreap J 3 19 24 15 ~ 5 Preah Vjhear Stueng Treng Battambang Pailin Pursat 23 21 20 Koh Kong Kampong Speu kampot Kep Takeo Svay Rieng 8! Kandal 14 Prey Veng T6fRatlnaFKfri HjMondul Kiri 10 Kratie 3.Kampong Cham 12 Phnom Penh 18jSihanoukviiie 61 Kampong thorn 4 Kampong Chhnang 12.3 27.8 2778 13.9 13."9 8.9; "8.9J 38.9] 20.0 277j 11.1 11.1 % 13.2! 16.0: 17.9 7.6 9.5 9.5 13.9 16.8 17.2 11.1 13.4: 19.8 Violence- conflicts over land, theft and robbery, and violence in home (Commune, aggregated up from village) GIS CODE (COMMUNE Insecurities Conflicts Over Land Violence in Home Total 160603 lAekakpheap 0 1 4 5 40801 Akphivoadth 13 12 27 52 80601 Akreiy Ksatr 18 5 45 68 50801 Amleang 3 8 26 37 30601 Ampil 37 126 " 1 0 3 266 141201 Ampil Krau 1 7 40 48 20405 Ampil Pram Dae 6 34 f 38 78 31101 Ampil Ta Pok 30 20 37 87 40501 Ampil Tuek 11 44 35 90 80101 Ampov Prey 0 26 62 88 20702 Andaeuk Haeb 2 2 11 15 70804 Andoung Khmaer 19 24 16 59 60112 Andoung Pou 14 18 35 67 I 40601 Andoung Snay 10 40 I 53 103 180201 Andoung Thma 0 15 13 28 200403 Andoung Trabae 14 6 22 42 90101 Andoung Tuek 12 9 22 43 211006 Angk Kaev 3 2 15 20 211002 Angk Khnaor 2 8 12 22 70101 Angk Phnum Tou 2 6 17 25 50301 iAngk Popel 9 9 17 35 210401 iAngk Prasat 14 10 12 36 200404 Angk Prasrae 5 6 13 24 210901 Angk Ta Saom 10 40 48 98 200501 Angk Ta Sou 7 3 21 31! 210601 Angkanh 2 34 67 103! 230101 Angkaol 4 17 10 31! 140601 Angkor Angk 2 18 14 34! 30701 Angkor Ban 5! 4 11 20! ...1628 1 1 9 Political Security Human rights violations per 100,000 (Province) GIS C O D E P R O V I N C E rHR_VIOLATIONS Banteay Meanchey 44 Otdar Meanchey 0 [17 Siemreap 25 "13 Preah Vihear 6 f l 9 Stueng Treng 1- 0 *02 Battambang s— 3 7 r 24 Pailin 0 "15 Pursat 34 '59 Koh Kong 25 r 05 ~ " Kampong Speu 34 "07 Kampot 39 r23 Kep 0 Takeo 26 "20 Svay Rieng 0 •08 Kandai 0 [14 Prey Veng 461 [16 Ratanak Kiri 6 "ii Mondul Kiri Oj b ° . . . Kratie 0 TJ3 Kampong Cham 46 ? 2 Phnom Penh 94 f18 H Sihanouku'lie 0 *06 Kampong Thorn 62] •04 kampong Chhnang , 26] Environmental Security Flood - Affected People per 100,000 (District) Drought - Affected people per 100,000 (District) GIS CODE ! DISTRICT 1 Drought affected Flood affected 205;Aek Phnum 16011 Andoung Meas 8081 Angk Snuol 6306 21011 Angkor Borei 1126 701 i Angkor Chey 4050 17011 Angkor Chum 1205 i702! Angkor Thum 580! 2201 Anlong Veaeng 504 Aoral i 1401 !Ba Phnum 826 1537 1501 Bokan j . .J252 16021 Ban Lung 2oTlBanan |- 10SU7 22021 Banteay Ampil 702! Banteay Meas 4937 3702 i 1703!Banteay Srei 44i 1603 Bar Kaev 601!Baray 401 jBaribour 50l'_Basedth 301 Bathoay 2102 Rati . 7 7 1 ™ " ~ . 5 4 8 0 5 " 1 6 5 8 3 S 7 7 1 ° 7 i ! 203 Battambany 204 Bawl 901 iBotum Sakor 2103 Uourei Cholsar 18705! 302iChamkar Leu 1201 iChamkar Mon 2001! Chant rea 2115 ... 185 Appendix 6 Base Maps and High Threat Maps for 13 Human Security Threats Economic and Food Security Health Security 121 122 125 Personal Security 130 Political Security Environmental Security 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items