|
|
|
|
Early Warning and Early Response, by Susanne Schmeidl and Howard Adelman (eds.)
What is early warning? Is early warning simply sharing information among those charged with responding to a conflict or a humanitarian crisis, so that each agency can determine what it can and should do to mitigate or prevent and impending disaster? Or is it simply a complex fire alarm system that signals that a pile of combustible material has burst into flame somewhere around the world with the inevitable production of corpses and hundreds of thousands of refugees in flight? Is early warning geared to action to tell when the station master at the UN in New York should dispatch the peace keepers? In another possibility, early warning is depicted as a complex system of indicators to predict the probability that a pile of combustible materials is likely to burst into flame. In that conception, early warning does not communicate what is happening, or signal that something has happened of crisis proportions, but that a crisis is likely to happen so that preventive action can be taken on time. In sum, is its task (1) to collect and share information on possible or impending humanitarian disasters or conflicts in others states; or (2) to sound an alarm; or (3) is it a complex set of indicators which could be used to predict, with some degree of probability, the type, timing and extent of disaster or conflict; or (4) a combination of the above?
Or again, is early warning neither just information sharing, nor signals system nor a predictor? Is it neither coterminous nor prior to a specific event, nor just the system for gathering and distributing information, but entails analyzing that information and developing response scenarios related to analysis at any point in time in the life cycle of complex humanitarian crisis? Early warning is then a monitoring process of a dynamic series of events in a crisis area. In that definition, early warning is directly akin to intelligence analysis, except that the focus of that analysis is not threats to oneself so much as threats to others who will be in dire need.
The multiple definition have created confusion around early warning. In many ways there seems to be little consensus on what early warning is, what it should do and to whom it should be applied. The confusion is multiplied when one finds that proposed indicator systems have been multiplying at a faster rate than any ability to conceive a test, let alone subject any of the indicators systems to such a test in order to measure the efficacy of each of the indicator systems proposed. This conceptual and methodological confusion fits into a predisposition by a number of political actors who have a propensity to believe that the issue in preventing and mitigating humanitarian disasters is not any lack of intelligence, but the lack of political will of states to involve themselves in conflicts not vital to their state interest. We already have too much information, they declare. What we need is the will to act upon it.
With such a predisposition, the conceptual confusion and methodological muddiness has led to the discrediting of early warning among a number of policy makers. This volume is dedicated to making the argument that it is the conception of political will that is confused, and to clarifying the misunderstandings of what early waning can and cannot achieve related to and as a precondition of developing political will. For without information and intelligent analysis, political actors cannot develop reasonable strategies for intervention. Proposals for doing so appear ad hoc and are the main reason for any lack of will to do anything. If you do not have a plan of action that appears reasonable and likely to be effective in advance, the chance of engaging in any humanitarian intervention is greatly reduced, as is the probability of such intervention being botched if it is launched.
The purpose of this book is to place early warning under a critical lens. Though there have been a number of workshops over the past decade on different aspects of early warning, the articles in this volume have been culled from the very first large comprehensive international conference ever held on early waning in March of 1997 in Toronto. In four different sections major questions and debates around early warning are tackled. Specifically, we have selected articles that deal with the conceptual meaning, the institutional locale, and the methodological relationship to the user side of early warning for purposes of response.
The book has two clear objectives. The first is to paint a clear and comprehensive view of early warning to end the conceptual confusion — though not methodological debate over which system of indicators is best when that approach to early warning is utilized. By clarifying the objectives of early waning, the links to the response side will also be clarified enabling policy makers and observers to understand where early warning fits into the tool box available for international actors dealing with areas of potential and actual humanitarian crisis.
Introduction to Part I: Preconditions to Early Warning or Why?
Why undertake early warning? Was there not plenty of information around to indicate that a huge massacre was likely in Rwanda in 1994? Was there not plenty of information to indicate that if the Hutu refugees in Zaire were not repatriated and the innocent separated from the refugee warriors of the ex-FAR and interhamwe who perpetrated the genocide in Rwanda, that the situation in Zaire would explode? Why make the effort to institutionalize early warning?
do anything. If you do not haveThe chapters in this section tackle that question only indirectly. For they ask a prior question. What are the preconditions for undertaking early warning at all? In the first chapter of this section, the definition of early warning is not debated; it is simply taken to mean the collection of communication of information, the analysis of that information, and the construction of scenarios for response to humanitarian crises and conflicts where the responders themselves are not under threat. It is the latter factor that makes early warning different from intelligence. If the agents, from whom a preventive or mitigating response is expected, are not under threat, if the action demanded is a humanitarian one even if it is one of humanitarian military intervention, what interest would any nation have to respond let alone gather any information, undertake any analysis, and develop scenarios for responding to a crisis?
Asserting that both responding to and the preparation for response should be done for humanitarian reason begs the question. The radical contrast between selfless motives and self-interested ones does not stand up, even though the fundamental purpose of compassion is to undercut the primacy of cruelty, murder and mayhem, the worst of self-centered acts. For acts done to ward off threats and acts of compassion, the chapter argues, both entail the motive of pride, the desire to make a statement about who we are and what we stand for, and the value for which we are recognized by others in the world. At the same time, the motive to act for positive reasons to express values is low in comparison to the motivations to act when those same values are under threat. Given the necessity of coordinating the responses of many actors governed by the same compassionate motives (perhaps to ensure that the accumulation of low motives can total the same motive of a singular actor under threat), creating a common information and analytical basis for choice is even more critical in humanitarian actions than in the intelligence works required to ward off threats, potential or actual. Otherwise, the same pride will convert an international humanitarian action into a display of ego and competitiveness that will reduce any such action to the lowest common denominator if it does not castrate it altogether.
Thus, the chapter argues that early waning is not simply a useful tool for humanitarian action, but the sine qua non without which effective humanitarian action cannot be undertaken. Further, continual contact between those charged with the analysis and development of scenarios for response, and those charged with making the decision is critical in order to build up trust to offset the destructive effects of unbounded pride. Finally, the early waning analysis must include assessing the comparative capacities of the different responders, including the pride and interest in taking ownership and leadership of the humanitarian enterprise.
Chapter 2 uses the case of the shooting spree of an Israeli doctor who killed 30 men at prayer in a Hebron mosque. The case entails action by a state concerned with both its own security from members of a group that threaten that security, as well as a humanitarian motive in protecting the lives of those same people when they are doing nothing that poses a threat, quite aside from the threat that the failure to prevent such an action posed to both the peace process in which Israel was engaged, the potential for a wider religious war, and the values to which Israel is ostensibly committed. The issue of early warning was at the heart of a commission of inquiry into the matter. Like the first chapter, the case is made by the commission, and upheld by the author, that early warning is indispensable for any effective response. Though a necessary condition, it is insufficient in itself and is no substitute for good strategies or shortcomings in the ability to respond.
But there were two contradictions. Those undertaking early warning analysis for intelligence purposes insisted that early warning was about types of events and the probabilities of risk in general, whereas those charged with preventive actions wanted and needed specific predictions. Secondly, routine training, so necessary for effective responses, also makes one unfit to anticipate atypical situations and uncharacteristic events such as the massacre in Hebron (or the genocide in Rwanda for that matter). At the same time, early warning analysis (and intelligence analysis) is normally best dealing with repetitions of what is similar rather than the novel. Thus, both the training given to responders and the inherent character of early warning to build on extrapolations from past experience are handicapped both in understanding and responding to novel situations. But if early waning is generally about uncharacteristic or atypical events rather than repetitive ones, then lateral thinking, imagination and creativity, rather than deductive reasoning, are the prerequisites for both early warning and effective responses.
If the first two chapters stress affects (pride) and imagination rather than deductive reasoning as prerequisite for early warning analysis and intervention, chapter 3 explores a third paradox of early warning — that it can never prove its own success. Further, it begins with another but more general and complementary non-rational factor in early warning, the risk averse propensities of decision makers (which the author calls decision takers), an aversion determined not by a rational assessment in the crisis area, but in the non-rational factors in his own bureaucratic and political situation. Further, although knowledge is a prerequisite to any decision making, the kind of knowledge we gather to evaluate decisions reinforces risk-averse behavior. Still further, the evaluations of that behavior, based on result-based management, reinforces the same propensity by only measuring results in conformity to prior policies and decisions rather than the relevance of the decisions themselves. Finally, the public nature of decision making in democracies, and the current interest of the public in the foreign policy field, in turn reinforce the same results based evaluations, risk averse and routinized behavior, all of which are unable to deal creatively with a crisis at an early stage and instead restrict development aid to measures of whether that aid achieves the objectives the monies were allocated for, rather than contributing to the overall objective of development of the country per se, including ensuring that a situation does not move into a crisis or a conflict.
Chapter 3 then makes a much larger, but complementary argument to chapters 1 and 2 — namely that unless we understand how we have institutionalized processes which favor routine thinking over creativity, technocratic measures over craftsmanship in the design of program, and evaluation processes which in turn reinforce those same propensities, the kind of information we develop and the people for whom we make it available will not contribute one lota to conflict management or even early warning when what is needed is a process which requires very different types of feedback in evaluating past actions, a very different type of people who undertake those evaluations, a very different type of people to whom those evaluations are provided and who are vested with the responsibility for using development aid funds for the benefit of a country, and finally, a very different system which ensure that the users of such aid have a mode of information collection and analysis available which enables the countries themselves to create a solid basis for crisis as well as conflict prevention.
Chapter 4 undertakes an analysis of the UN to demonstrate why it is inherently ill-equipped to undertake the analysis and scenario-building requisite to early warning. For the UN is not only the most risk-averse institution, not only prone to a narrow basis for evacuations concerned with ensuring honesty and conformity to routines in the use of monies rather than creative action in the prevention of a crisis in the first place, inhibited from any actions that can be perceived as an intervention in the sovereign authority of a state, and political restricted in creative interventions in intra-state conflicts before they reach a crisis and overt conflict state, but given the international political context and the structure of the Security Council, would seem to be incapable of undergoing the necessary reforms to make the UN the center for any early warning system.
However, chapter 4 suggest an answer which allows the inclusion of the UN without depending on it to be at the center, allows the circumventing of technocratic managed bureaucracies in states without requiring as a prerequisite that they be reformed — that is the construction of a decentralized Internet-based early warning system — much along the lines proposed by the Forum on Early Warning and Early Response (FEWER). Unfortunately, chapter 4 does not answer the question of how to overcome information overload, how to develop the trust between the analysts and the key continual contact, and how to allow triage to be exercised given the multitude of crises and the multiplicity of analyses and solutions inherent in such a decentralized system, and how to institute coherence when the anarchy and incoherence have been cited as central factors reinforcing impotent interventionist responses. If chapter 1 is correct, then ego and competitiveness would be reinforced in such a system at the same time as the ability to create pride ownership and responsibility would be undercut. For some answers to these questions, the reader will have to examine the chapters in the subsequent parts of this volume.
Introduction to Part II: The Information Gatherers or Who?
One of the most frequently asked questions, aside from what early warning is, tends to be ‘who should be responsible for early warning.’ This question is actually split into two parts, almost a double-edged sword, who will provide the early warning and who will act upon it. Yet, to further complicate matters, one could ask, who provides the information for the analysis that then leads to policy recommendations for policy actors. This implies, that early warning may possibly be a division of labor among various actors. Again, this multifaceted view of early warning reflects back on why several definitions exist.
While it might be of interest to discuss the final actor, it seems there is more consensus than disagreement in this area. Ultimately it looks as if governments (often powerful ones or an alliance of multiple governments) react in cooperation with the United Nations (which alone might be too weak). Joint-action and UN sponsorship eases the trouble of intervening in the sovereignty of other governments (as we did in Iraq). Yet even if there was confusion on who should respond to early warning, it seems discussing this would put the cart in front of the horse, such as that without proper early warning, no response can be expected.
Since the analysis part is tackled in the next section, it seems most interesting to focus on the information gathering as the task to be accomplished and by whom. While many may perceive the collection of information as neutral and innocent, in political laden environments it can be very dangerous (protection of sources) and powerful (influencing responses). Thus, the question is where to obtain neutral, impartial and credible information for early warning purposes that does not try to gear analysis (and response) into one direction. After all we would want to avoid that information jeopardizes the weak and vulnerable in a political unstable environment. Furthermore, we would also want to avoid that the country in question hides behind international (or US) conspiracy claims when available information sends out a warning signal (as Iraq often takes the stance). The answer might be to rely on multiple sources in order to compare information and verify its accuracy.
This exploration of multiple sources is exactly what the chapters in this section undertake. They survey the different potentials of information gatherers from NGOs, academic networks, the United Nations, to the media. The basic question asked is not only how do those actors collect and analyze information, but to what extent can they be reliable providers of early warning information.
Chapter 5 discusses the importance of information to early warning in general and the author concludes that a new integrated global approach is need for linking early warning with action. Her metaphor is striking in that information without analysis is like an orange without sunshine and one could go on further to argue that without the link to action the orange will not be eaten. Sharon Rusu continues to argue that no single organization can do the task of early warning and information gathering and a general cooperation is necessary. Of course, while we need to share information, we need to agree on standards to increase credibility, as well as focus on quick access to make early warning indeed early and not late. As an example of such a coordination body of information, she discusses ReliefWeb, a global information system developed and maintained by the UN Department of Humanitarian Affairs. ReliefWeb was a response by the United Nations to the impasse of information flow during the 1994 Rwanda genocide. The idea is to provide a platform for information dissemination using the world wide web to serve the UN system and others. The chapter discusses how ReliefWeb can contribute to existing early warning efforts and how it can be useful. In a very reflective and self-critical manner, the author acknowledges the fact that ReliefWeb is merely a tool that can contribute to early warning analysis, but in-itself does not lead to preventive diplomacy. However, one can only applaud the effort of ReliefWeb which has an increasing amount of users every day. What it shows us is that we are able to collect information and disseminate it fast. Thus, policy maker cannot claim that a lack of action was based on a lack of information.
Chapter 6 asks the questions whether NGOs can play a role in early warning and the collection of information for that purpose using the example of the genocide in Rwanda. The focus on NGOs arises from the authors observation that NGOs seem to have some structural advantages (size, network, non-governmental) over other organizations in the field. However, examining the case evidence, the authors conclude that NGOs did not make use of their advantages and capacities in the field and thus failed in providing adequate early warning. The authors attribute this to the fact that the large scale RPF offensive in 1993 led to internal displacement and thus a new humanitarian need that needed attention. In the course of humanitarian assistance, NGOs quantitatively increased in their service but decreased in their qualitative service due to a larger geographic spread, new field staff and simply an overall exhaustion. Thus, when NGOs needed to be more sensitive, they actually lost sensitivity for developments on the ground. The authors then conclude that in order to learn from the Rwanda experience and work toward early warning capacities in the future, NGOs would need to maintain and deepen their connection with local communities despite the energy drain any emergency can pose to an organization and its leadership. In addition, NGOs need to improve their understanding on how to use their assets and capacities on policy and operational levels in order to feed into an international process of early warning.
Chapter 7 and Chapter 8 tackle the utility of media in early warning which “suffered” some credibility by not only failing to predict the 1994 Rwanda genocide, but even failing to recognize its occurrence by initially labeling it a mere ethnic conflict. Chapter 7 looks at print media and electronic wires while chapter 8 analyzes television coverage. While not covered in either article it is worth to note that media (Radio Mille Coline) can be (and frequently has been) used to spread genocidal messages. This, of course, confirms some of our suspicions regarding the sensational character of media and its lack of credibility. Nevertheless, many people (including policy makers) still use it as a source of reliable information. In addition, media can provide the necessary pressure for the international community to get involved into an ongoing conflict (see Somalia). Therefore, both chapters evaluate to what extend media can be a credible source of early warning information. In addition, the chapters examine if media can live up to our expectations by actually predicting humanitarian disasters and also impacting on international responses to such crises. Unfortunately, both chapters seem to reach the same conclusion which is that media might have a tremendous impact on the general public, but it most certainly cannot be used for early warning purposes.
Chapter 7 analyzes the coverage of Reuters and New York Times data of the first three years of the Palestinian intifada to measure potential change in media coverage throughout the development of a crisis. One big question obviously is the issue “media fatigue” and the importance of competing news events. Their finding is that albeit Israeli–Palestinian relations are generally covered quite well by media, the coverage indeed dies down as longer as a conflict continues. Thus, the find support for media fatigue, but also for the fact that some of this decline is due to competing stories becoming more important. Even more importantly, the authors analyze whether the fact that a story got reported at all (or an area is covered by journalists) is more important than the actual content of such story. This would be important for long-term monitoring using machine or human coding of news paper events. The findings are a bit sobering since the analysis suggests that the detailed coverage of a story itself seems to have little impact, boiling down to a coverage vs. no coverage of an event. Therefore, the authors conclude, more serious early warning analysis might benefit more from specialized reports as found on ReliefWeb than generalized media articles.
Chapter 8 uses American television networks (such as CNN) and their coverage of Rwanda to assess if they can serve as a response catalyst in humanitarian crises. Unfortunately, the findings are negative again. The authors, however, dig deeper into the problem of why media fails to cover certain events. The article suggests that we should not expect too much from the media since particularly television networks are consumer-dependent and therefore may cover less important events. Thus, besides looking at the coverage of events, the article also analyzes the reasons as to why Rwanda was not covered as it should have been. First, there are obvious competing news stories (as already pointed out in the previous chapter) that seem to be more tailored to the interest of the general public (or closer to home) than genocide in far away lands with backgrounds that are hard to grasp. The example given is the heavy coverage of the O.J. Simpson trial during the Rwanda (and Bosnia) genocide. In addition, there is almost a chain leading up to lacking coverage in certain regions. The argument is that the less a region was covered in the past, the harder it will be for journalists to explain to the public why a genocide might be happening. Given that consumers often want simple stories they can understand, complex backgrounds to a crises in itself might censor the coverage. And, there is, unfortunately the fact that journalists are often too new to the area (such as Africa) and thus might lack important background information themselves to even understand an emerging crises. Despite all the criticism, the authors do however defend journalists by arguing that it sometimes is lack of access in dangerous situation that negatively affects news coverage. In sum, the article concludes while the media might aid in raising public awareness, it is far from working in an early warning network where information is analyzed for the purpose of detailed policy recommendations that can lead to preventive action.
The last chapter in this section examines the potential of military and civilian observers that are used by the United Nations and regional organizations during the various stages of escalation and de-escalation of conflict. The author concludes that as currently utilized, observer missions tend to produce more late-warning information since they are often deployed after violent conflict seems very eminent or has already started. In addition, observer missions are often limited in their mobility in a country and also in their time to collect information. Furthermore, so far they have not been trained to systematically collect information and link it to the emergence and spread of conflicts. However, the author sees great potential that once observer missions are better structured, trained (in data collection and analysis), and deployed to address the many dimensions and incipient violence, they can be used for early warning purposes to ignite international response. Nevertheless, an appropriate deployment, increased mobility, adequate structure and a linkage of those missions to an effective early warning system (or mechanisms) or access to policy makers seems a necessary precondition for success.
In light of the above, while information is certainly a necessary prerequisite for early warning, it does not seem to lead by itself to a response by policy makers. In a consumer oriented world, we may assume that policy makers have the ability to chose from the products (information) that exist. But, in the age of information revolution and the Internet, there is almost too much information available which can overwhelm consumers as to how to select the appropriate pieces and also as to how to react to it. Thus, we need to package the information correctly, or we need to analyze it and present suggestions as how to interpret is. This leads to the contributions in the next section.
Introduction to Section III: Quantitative Approaches to Early Warning and Risk Assessments or How?
In the field of early warning and early response there has been much debate on the correct methodology of early warning. Traditionally policy analysis is inductive, working with specific information in any given event and using experience to make sense of it. The quality of such analysis, however, then depends on the individual experience and the ability to related past actions to new developments. In an effort to systematize early warning analysis, some academics have proposed to use a deductive approach or quantitative assessment of information in order learn from past events and crisis. This means, what we learn from past conflicts can be used to predict new conflicts using econometric forecasting and predictive models. Given that such analysis differs fundamentally from policy analysis, and to some degree necessitates an understanding of quantitative methodology and its interpretation, it has often been criticized and even marginalized among those interested in early warning. Furthermore, the fact that there are multiple methods that can be used for predictive analysis has further confused those not familiar with the methods. Therefore, this section has selected papers that have begun to evaluate their methodologies, and show its use and application for early warning. The goal is to show what quantitative modeling (often also referred to as indicator approach) can contribute to early warning and how it can be made accessible to policy makers that may not have much time to cut through a large array of formulas and evaluate one method over the other.
Chapter 10 provides an overview on the vast methodologies that exists and could be utilized for early warning activities. The author surveys available econometric methods that are used to forecast events by examining the closeness of fit between the predicted and the observed value. Despite obvious questions on the utility of the various methods of analysis or the quality of the data used, the main point the author tries to bring across is that we need to move away from the idea of forecasting toward the mere idea of ‘warning.’ This argument is based on the fact that in early warning, predicted values (those that we wish to forecast) are not independent of the observed value (the actual crisis), a violation of a major assumption of econometric methods. But, we should not desire anything different, since in essence this link between predicted and observed value is almost wanted in that we would never actually have a crisis occur after an early warning was sounded. Similarly to Chapter 4 by Jean Guilmette, Gupta reminds us that the best early warning system will never ever achieves its predictions, thus always prove. This is of course what we want, we want to predict the outbreak of a crises, with the hope that early action will prevent exactly such crises.
Another a major problem of humanitarian early warning is exactly the human factor, thus even if we were able to create the perfect model, we would have a hard time predicting the unpredictable, human action (or non-action). In sum, the article clearly lays out what quantitative modeling can achieve and what it cannot. It can sound a warning to policy makers that certain countries and regions have all the elements of a future crisis, it might even be able to rank the countries into a hierarchy of risk, yet it will never be able to predict for sure if a conflict prone situation will erupt and when.
Chapter 11 also starts with a similar premise in that it makes a difference between early warning and risk assessments. They would argue that empirical-derived risk assessments can identify situations with a high potential for future conflict, where then preventive action might be necessary, will never be able to accurately predict the actual outbreak of such an event. A major reason for this is that most quantitative models assume variables or conditions to remain constant, in that we predict a certain outcome if the underlying factors do not change. Similarly as Gupta, the authors would agree that obviously preventive action will change the conditions for conflict and thus conflict potential may decrease. In a second step of the theoretical discussion of comparing different methodologies, the authors predict ethno-rebellion in the mid-range future, using three different models that are based on a different set of analytical assumptions. The findings of the analysis are then compared against each other and against case evidence. The very interesting point of this paper is that the three methods identify similar but not the same groups with a high risk potential, with a few exceptions of overlap. Yet the similarities are great enough to assume a not fully random selection process by the methods and one could start to monitor the few groups that were predicted by all three methods. In conclusion the authors argue that multiple approaches of risk assessment should be encouraged for comparative purposes but trying to pick the one best method is clearly a waste of time. In addition they try to emphasize the importance of quantitative methodology in providing theoretical interpretations for future outcomes and identify cases worth monitoring.
Chapter 12 now uses the idea of economic forecasting for early warning purposes. While Gupta had suggested that we should not focus so much on forecasting, Rowlands tries to show test the effectiveness of different forecasting models. The author provides a spin on early warning by moving backward from predicting the actual conflict to the root causes of potential conflict—economic instabilities. By analyzing the impact of IMF emergency lending on recipient countries the author opens the door to actually evaluating a potential factor that could prevent conflict if applied properly. Aside from discussing differences in prediction when using different time lags, methodological assumptions and dependent variables, the results seem to suggest that a short-term forecasting in the economic realm is quite possible. While obviously much can be improved, this article is a good example of very long term early warning addressing root causes as compared to the problems of short-term forecasting of political conflict.
At the end of this section, however, it still appears that we are missing yet another piece of the early warning puzzle. We are now able to obtain information from various sources and in many cases fast enough for actual early warning. The section on analysis has shown us that we have progressed to a point where we have the necessary tool to make sense of information and more importantly learn from past conflicts. Nevertheless, it seems we still only have a product. At this point the product might be packaged correctly and ready for sale, yet are we able to get the consumer to actually buy it? It seems as if early warning analysts need to do marketing research in order to find out how to best bring their product (early warning information and analysis) to the consumer, so that s/he will actually make use of it. This link between early warning and response has been long ignored in early warning literature which had an heavy focus on analytical issues.
This is the problem that the last section of this book deals with. We now will close the circle that lead us from placing early warning into context, discussing the gathering of information, making sense of it with analytical tools to ultimately the selling of the product to policy makers. Only now that there is an adequate response to early warning can we argue that it actually worked.
Introduction to Section IV: Early Warning and (Early) Responses or What For?
The quest for a humanitarian early warning system began with the initiative of the FAO in creating the Global Information and Early Warning System (GIEWS) to monitor the exiting and prospective global food supply in relationship to demand in order to anticipate where shortages might develop and to mobilize donors to position reserves in order to prevent famine. Early warning was created in order to be able to carry out a specific response much more effectively. What is interesting to note is that even in this relatively uncomplex situation, where the measures are direct and can be forecast with relative accuracy, where the responses have the greatest degree of uncontamination from political issues and are most purely humanitarian — assuming that there is such thing — even then there is a gap between the early warning information and the political will to deliver a response that is both adequate and timely. Yet early waning continues to be the sine qua non of whatever success has been achieve, and in famine early warning that success had been considerable.
The above reference to a ‘relatively uncomplex situation’ does not mean that assessing the vulnerability of specific population groups does not entail measuring the relative weights of a host of complex and multi-correlated factors, but only that, relative to conflict situation, there is complexity and there is COMPLEXITY. Further, since many of these famine situations arise when civil strife or war hamper the distribution of food, the two types of crises overlap. To demonstrate the effectiveness of early warning in famine relief, where early waning proved to be a necessary but insufficient condition for ensuring that such relief was available, it is appropriate that two examples were chosen based on ‘natural’ disaster induced famines — namely from drought in the sub-region of Southern Africa and floods in North Korea, one in which the early warnings were effective in enabling assistance to be provided in a timely and sufficient manner, and one in which an adequate response did not follow directly from the adequate early warning. In the latter case, the causes were also more complicated since the organization of the economic sector in North Korea led to increased reliance on the production of fertilizers and the import of essentials, such as chemicals and spare parts, just when productivity had been declining and the financial ability to import had fallen. These same political economic factors inhibited donors from responding adequately and quickly lest the donations be used to strengthen what is consider to be a rogue regime. What may also be clear is that FAO’s early warning system is limited by not analyzing the political factors that limit responses as part of the early warning system, a limitation which is at once understandable but also restrictive.
In contrast to chapter 13, chapter 14 provides a case study on how early warning is done in a humanitarian organization—the UN—on humanitarian disasters. Obviously there might be similar issues, but humanitarian crises include the unpredictable human factor. Nevertheless it is interesting to see the two contrasted. In chapter 14 Walter Dorn surveys the constraints and opportunities the Secretary General has for early warning and early action and also makes suggestions on how such obstacles might eventually be overcome. Unlike Duffy in the first section of this book, Dorn is a strong supporter of and early warning within the UN, but one should take his suggestions with Duffy’s criticism in mind, to put it into a larger context.
If chapter 13 and 14 provide an argument for both the necessity of early warning and its limitations when ideologies and politics are involved, chapter 15 suggest another reason that early warning has not been taken seriously — namely that it is not clear what would be the best thing to do once an early warning was received. The chapter tries to provide a set of tools that can narrow the gap between the early warning — assuming that an adequate system is in place — and an adequate and timely response. It does so by using the analysis component of early waning combined with empirical evaluation on past responses as a foundation to derive the most appropriate response mechanism. The judgment that early warning is superfluous because there is a lack of political will may simply mean that people did not know what to do that would be effective, not that they were unwilling to act. Perhaps no one could agree on a way to respond. The analysis does reinforce dispensing with the ‘alarm’ idea of early warning and strengthening the conception of early warning propagated in this volume — that is a conception of early warning as the collection, communication, and analysis of that information, along with the development of appropriate responses.
This Chapter indirectly raises another problem — the culture gap between those who undertake analysis and those charged with the responsibility for making decisions — another factor which makes decision makers wary of the value of early warning. As Lund states it, the responsibility for change must come from analysts concerned with preventive action engaging in more studies on the effectiveness of different tools at different stages in a conflict cycle, while decision makers must be more open to early warning if they want to minimize ineffectiveness once they decide to take action.
Chapter 16 makes precisely the same point as the previous chapter — that we lack the knowledge to determine not only which actions taken are effective, but who should take them and when. Even with respect to the use of deterrence in the form of coercive diplomacy, the most widely researched of the intervention tools, the evidence is ambiguous. Carment and Havery attempt, however, not only to point to the gap but to fill it in some measurable degree by assessing the success rate of third party interventions related to different kinds of crises. The answer reinforces the current predisposition to rely on multilateral interventions rather than either unilateral actions at one end of the spectrum or UN led actions at the other extreme. Further, it also reinforces the position that has been a constant theme throughout this volume, that humanitarian early warning — what authors deem analysis, trend assessment and political forecasting in the absence of threat — is critical to effective and timely response.
However, it is fitting that the final chapter end on a much more skeptical note — arguing that the causes of both the conflicts and the inability to respond effectively are rooted in similar and much deeper causes than heretofore discussed. The chapter implies that early warning cannot be effective in any fundamental way at all under existing circumstances because the underlying staple of intra-state conflicts, but in the inter-state system itself, and the past and current actions of states in that system. Instead of simply concentrating on early warning and effective responses based on them, the authors propose a total overhaul of the nation-state system with a focus on what they call ethno-development. Only integral to such a development do they propose that early warning be developed based first on a common thesaurus, then on exchanging and accumulating data on ethnic conflicts, and finally linking that data to effective responses.
| Susanne Schmeidl and Howard Adelman September 1, 1998 |
Email this citation