Early Warning and Early Response, by Susanne Schmeidl and Howard Adelman (eds.)

 

15. Bridging the Gap between Warning and Response:
Approaches to Analyzing Effective Preventive Interventions ₁

Michael S. Lund
Center for Strategic and International Studies
and Creative Associates International, Inc., Washington,D.C.

Early warning data systems have been proliferating for some years and gaining verification. But analysis of what individual and multiple responses are effective in preventing violent conflicts has begun only recently. Until high-level policymakers have received authoritative analyses of what preventive interventions work under what circumstances, we cannot say whether their failures to respond arise from lack of will or lack of a way. The knowledge policymakers can use involves several key action-oriented questions—when? what? why? who? how? and whether?—but existing early warning research appears to address many of these questions only indirectly or tangentially. The paper raises the issue of what units of analysis might be most useful for codifying recent preventive experience, and describes recent work using two forms of such effectiveness analysis. These forms are case-studies of actual recent successful and unsuccessful multi-tooled preventive interventions (e.g., Macedonia, Burundi), and evaluations of differing policy tools of intervention (e.g., preventive deployment, conditional development aid). How this knowledge might be incorporated into the country-level strategic plans of donors and other third parties is discussed. The resulting idea of rolling prevention, rather than the prevailing alarm bell model, is offered. Evidence and examples are drawn from the author’s and others’ case-study research on paired successes and failures in Africa and Central and Eastern Europe and Latin America, a guidebook with tools profiles done for the Greater Horn of Africa Initiative of U.S. A.I.D. and the State Department, and a review being done of what is known about development aid’s effect on conflict.

 

Introduction

The only practical raison d’etre of early warnings is to be better prepared for possible humanitarian crises and violent conflicts and to prevent them where possible. But though early warning research is well developed in terms of data and methodologically, very little research has been done on the best things to do once early warnings are received. While sideways synergy may be needed among early warning systems, Afront to back linkage between early warnings and possible responses are possibly more urgent. To paraphrase Fred Ikle’s January, 1961 Foreign Affairs article about arms control, After Detection, What?: After Warning, What? Just as he argued arms disarmers needed to be concerned about whether effective enforcement actions would follow after a violation is detected, so early warners who urge policymakers to pay more attention need to show that effective actions are possible. In fact, perhaps a major part of the reason that the world leaders who have subscribed to preventive diplomacy since 1992 have failed, however, to respond at critical moments to early signs of major upheavals is not simply the much-heard explanation that they lacked political will. They may not yet know what it is they can actually do and thus lack confidence they can make a difference. But if they saw a way, perhaps there would be more will.

To stimulate more consideration to developing effective responses to warnings, this paper starts at the other end of the desired sequential chain—when and where preventive responses take place—and works backwards. Useful insights might be gained if this process met up in the middle with the early warning work coming from the other direction. The paper asks:

• What knowledge do would-be preventive policymakers, in their station, need for responding effectively to warnings ?
• How much does most early warning work provide that knowledge?
• What are some ways to fill the evident gap?

 

What knowledge do early responders need?

Let’s imagine we are a decision maker with some authority to take preventive action, or an advisor to one, working in the UNDPA, in Ottawa at the Department of Foreign Affairs, the State Department in Washington, in Brussels, the OAU in Addis Ababa, or the country-level aid mission of a donor government or multilateral. We have information considered reliable suggesting the strong possibility of an outbreak of civil war or a breakdown of authority in a country of our concern. We have been briefed on the various sources of this potential conflict and have probed dutifully to grasp the details and nuances of the situation. Setting aside for this purpose the organizational and other factors that would bound our rationality, what kinds of policy and implementation questions would we want to have answered in deciding how to react? Put in simple form, they seem to be:

• What kinds of action(s) might work to prevent the conflict?
• When should action be taken?
• Who should act (within my organization and/or by other actors)? What direct or supportive roles should be played by whom?
• Why take action? What are its goals?
• How would this action be organized and implemented?
• Whether action should be taken. Should any response be made (as weighed in terms of interests or morally, and practical and political feasibility)

While knowledge is not the only factor that would inform the answers, it would usually be seen as valuable in reaching intelligent ones.

 

What response knowledge do we have?

Which questions would early warning analysis, with its conflict models and sequences, be able to help answer? Here are some impressions based on a quick survey. The main extant forms of that genre might be fairly helpful in providing a sense of the when to act. But few answers would seem immediately forthcoming about the what, who, why, how and whether questions that policymakers might be asking.

True, it might bear somewhat on the what and who. But what would it yield? Most early warning has focussed on identifying the multi-leveled causes and temporal antecedents of various bad events, so they can be better predicted and avoided in the future—we will see them coming. The general nature of some particular responses might be implied or inferred from such predicted outcomes: Possible genocide? Stop it forcefully. Somehow. But neither stating this possible imminent conflict nor intensive analysis of its systemic, proximate and immediate causes and dynamics, do not point to a particular solution, automatically generate a range of options, or demonstrate their effectiveness. Even had we been able to predict Albania’s rapid descent into near anarchy and collapsing regime legitimacy from the popular outrage, would that have advised strengthening security, pressing for new elections, or what?

However, other extant early warning approaches are beginning to track good events—cooperative actions and early stage engagements by protagonists or third parties that led to downturns in tensions or other peaceful directions. These are more helpful, for they at least note counter-actions that prevention-minded parties might replicate in future similar situations. But the aim seems still mainly to predict, not advise. For one thing, the policy implications of these retrospective chronologies are not generally elaborated or written up in usable forms for decisionmakers. Also, the conclusions are bound to the responses that happened to be made in those past instances recorded, little description is provided in operational terms regarding the nature of the preventive responses or why they were taken, the contextual conditions in which they worked are not examined, and little is developed about the impacts on dimensions of conflict beyond that of some bare vector registering hostility or comity.

Why does this gap exist? In part, because of some sociology of knowledge and recent history:

• By and large, early warning is done by social scientists, and social scientists have seen their role to be diagnosing the causes and nature of problems, not devising solutions.
• The incentives of advancement in universities are primarily to get it right, not to get it done. Thus, the dedication to ultimate rigor. Heuristic is not an honored adjective.
• On their part, government policymakers see the task of deciding what to do as their prerogative, thus leaving knowledge-building for policy to subordinant in-house analysts. But neither the former nor the latter have much time to harvest the best from the universities. Thus, a split has developed: conflict analysis and warning means pure knowledge; conflict response means pure action, and advocacy to spur it.
• Because of the Cold War, the knowledge that might be harvested from universities from such fields as conflict resolution and security studies (with a few notable exceptions such as Hugh Miall’s and Lincoln Bloomfield’s empirical work, frameworks and theories of conflict escalation dynamics by Louis Kriesberg, Ronald Fisher, and some others) has focussed almost exclusively on already-escalated crises or intensely violent levels of conflict, and post-conflict situations, and the tools that are most common at those levels (e.g., coercive diplomacy like sanctions, mediation, peacekeeping). But conditions and tools at the dispute or pre-violent stages and thus many tools useful and used in conflict prevention or preventive diplomacy, have not been much studied. Unless one thinks of tools like deterrence and arms control. But the latter typically have dealt with preventing potential inter-state wars, not the national conflicts more dominant now, in which the international community has gotten increasingly involved, but in a predominantly remedial way.
• The humanly compelling and more dramatic nature of humanitarian crises and wars rivets the media and everyone else mainly on killing and dying, and trying to stop them, not on preventing them. (If it bleeds, it leads.)
• Though major governments and multilateral organizations are heavily investing in alleviating recently-arising bloody national conflicts (e.g., Rwanda, the former Yugoslavia) as well as those left over from the Cold War period (e.g., Angola, El Salvador), they have not yet concluded that the destruction and bloodshed and their own financial and political costs have mounted sufficiently to warrant giving more than occasional time and support to systematic analysis of how best to prevent these violent conflicts from happening in the first place or to operationalizing decision procedures to devise responses based on such analysis. Yet, new funds are being set up and flowing to mitigate conflicts and for post-conflict rebuilding from the wreckage.
• The NGO’s which have often stepped into the breach of early conflict situations to play preventive diplomat are stretched thin across all kinds of conflict situations and stages and thus far, have not been able to do much to evaluate their efforts.

 

How can the response knowledge gap be filled?

In sum, for these and other reasons, a lacuna exists in policy analysis for prevention. A few analysts and collective projects decided a few years ago not to wait until early warning research had it completely right before trying to tackle systematically the ultimate policy question of what preventive responses work, why, and in what circumstances. One methodological issue that must be faced by such an endeavor and that bears on its possible utility for policymakers is the unit of analysis to examine: the project, the individual decision, the international intervention, what? Though few evaluations of individual projects or initiatives in fact seem to have been done, the upshot of single efforts in helping a whole vulnerable nation head off conflict is likely to be minimal, anyway. Thus, evaluation that seeks to assess impacts at a more aggregated level, however more difficult, seems the most fruitful. Two possibly approaches being pursued are:

• Comparative case-studies: Studying recent particular instances when deliberate preventive interventions were made into brewing national disputes, and they succeeded or failed to prevent escalation (e.g. Macedonia, Croatia, the Baltics, Guatemala in 1993, Haiti).
• Tools assessment: Evaluation of the strengths and weaknesses of individual generic policy tools of intervention used or possibly useful in many places (e.g., confidence-building measures, preventive deployment, military restructuring, track-two diplomacy).

Focussed comparisons of paired cases of success and failure

One way to structure comparative case-studies is to select pairs of countries in the same region that faced the same general systemic challenges and experienced resulting political disputes, but that differed in terms of whether these were handled violently or peacefully (e.g., Macedonia versus Bosnia, Estonia versus Moldova, Congo/Brazzaville versus Rwanda). The idea is to isolate the particular factors that account for the diverging courses of the conflicts and outcomes. The ingredients which are looked for are both endogenous within the arena of conflict and exogenous in the actions of third parties outside the region. A structured, focussed comparative approach is used to study the cases over similar periods of time. The initial conjectures as to which are most important variables to put in this framework and look for in various cases are obtained from extant hypotheses in relevant literatures, such as the causes of ethnic conflicts and escalation theory. Criteria for gauging observable or reasonably inferable preventive impacts on conflicts are defined in terms of levels of conflict intensity and qualities of peacefulness such as:

• containing low-level violence (violence prevention),
• engaging opposed political actors in communication and settlement (pre-emptive initiatives), and
• active, salient efforts to address structural disparities (preventive peacebuilding), and other such gradations.

This approach has several advantages for policy purposes:

• Looking at several cases rather than just one subjects the findings to more tests, thus strengthening the surviving hypotheses, and permitting enrichment of the explanations from one case to others.
• Looking at successes as well as failures allows examination of positive factors that were favorable influences in the structures, institutions and behaviors of those cases, and thus that can be identified as promising solutions, whether they are indigenous trends to promote by third parties, or effective actions by third parties themselves.
• Selecting pairs of countries with somewhat similar locales, histories, cultures, size, and basic time-specific systemic challenges helps in controlling for the influence of pervasive conditions that single case-studies often use to explain violent conflicts (e.g. , breakdown of the Soviet empire, economic liberalization), but that are simply latent pre-conditions, not determinative causes of manifest tensions or violence. This helps to expose the more contingent, proximate causal factors that actually emerged from the latent factors. And these proximate factors are of a kind that policy might have a chance to affect (e.g., political strengths of coalitions, timing of third party engagement, range of measures used).
• The findings about the most important variables from case to case can be used to construct a checklist of factors, or framework for risk assessment, used as alerts to trigger degrees of attention to prospective situations, in order to judge the relative severity or manageability of political disputes arising.
• The intensive case-study nature of the research permits a deeper understanding of the dynamics of escalation and de-escalation, the operational mechanics of the policy interventions tried, the specific causality through which they exercised their influence, and the gradations of their impacts.

These studies so far have identified a set of apparently key indigenous and third party variables whose particular character steers a country in a peaceful or violent direction. Briefly, these include, for example, the political autonomy of the state, the monopoly the state has over the means of force, the extent leaders demonstrate moderating words, deeds, and policies, as well as third party early engagement, multiplicity of tools, and major power unity. ₂ By describing the endogenous conditions, institutions and behaviors that have been favorable to conflict prevention, and thus that third parties can reinforce, as well as showing how effective third party efforts worked, the conclusions can yield useful insights for answering policymakers’ questions regarding when, what, who and how.

Preventive tools evaluative profiles

Another approach to answering response questions is to examine the strengths and weaknesses for prevention of generic policy tools that are used or can be used in pre-violence situations (e.g., traditional dispute resolution mechanisms, conditionality in aid, track-two diplomacy, war crimes tribunals, democracy-building, governing structures like autonomy, etc.). The array of tools used or potentially useful spans many functional policy sectors, and includes a large number of tools beyond the usual ones thought of.

The first task is to inventory this array, which can provide a long list (see the matrix attached). The next task is to review the existing evidence from case-studies and other analyses in order to profile their features and evaluate their performance. Thus, synopses can be developed that describe a particular tools operational features and typical sponsors, and that evaluate their performance in terms of such criteria as:

• implementability,
• political feasibility,
• interactive effects with other tools, as well as
• impacts under differing conditions, such as different stages of conflict (see attached first page of a sample tool profile).

Of course, some tools like mediation and economic sanctions have plentiful research literatures, while others are lucky if they have even been described in any detail, much less assessed in terms of their anti-conflict effects. Pulling together the extant evidence in a composite picture, and providing a set of such profiles to policymakers, creates a kind of toolbox of potential actions. This can have very practical value as a reference work or guidebook when the policymakers turn to considering which among an array of tools in place or recruitable might be useful in responding to specific situations. ₃ Among the most interesting findings that emerge from such composites is a set of generalizations concerning what kinds of policy tools are especially suited to different levels of hostility.

In sum , these two approaches—comparative case-studies of best practices as well as failed prevention in given sets of conflicts, and evaluative studies of generic tools for preventive intervention (both having endogenous and exogenous purviews)—seem to hold considerable promise for producing the kinds of if-then statements that policymakers can consider when faced with making particular decisions regarding the what, when, who, and how of contemplated preventive initiatives.

 

Toward rolling prevention

On an ongoing basis, the findings of case and tool studies can inform the periodic development of forward-looking strategic plans done at the country-level. Increasingly, aid administrators, donor governments, international lenders, diplomats and security specialists are thinking they need to take into account the utility or harms that their respective functional programs may be having in encouraging or discouraging national or local violent conflicts. Findings from both successes (such as Macedonia so far) and failure instances (such as Rwanda) suggest that the closer to the conflict arena that early warning, conflict analysis, and response decision making authority is—compared to sounding alarms from remote places and trying to mobilize the attention of distant, distracted capitals—the more ready the response and less problematic the issue of political will. Thus, inserting conflict prevention and preventive peace-building elements into existing diplomacy, development, trade and other ongoing relations—so small, repeated preventive peacebuilding measures can be taken to keep a nation on course—may be the optimal form of conflict prevention, if it can be achieved organizationally. ₄

Of course, the next step after considering the applicability of any case-study variables and tool profiles is to look closely at the peculiar features of each conflict situation—at the particular moments in time when they are being addressed—in order to see how changing and nuanced conditions correspond or depart from the generic patterns suggested, and to make the specific ireducible judgments about actions that the decision maker must ultimately make. But hopefully, systematic analysis of multiple and single interventions will help to inform the making of such judgments and bolster the policy maker’s will to do so.

 

---

Endotes

Note 1: Discussion paper prepared for conference, Synergy in Early Warning, sponsored by Centre for International and Strategic Studies, York University, Toronto, Canada, March 15–18, 1997. Please do not quote or cite without permission. Back.

Note 2: See, e.g. Michael Lund, “Preventive Diplomacy for Macedonia, 1992–1996: Containment Becomes Nation-Building,” chapter in Brue Jentleson, Preventive Diplomacy in the Post-Cold War World: Opportunities Missed, Opportunities Seized, and Lessons to Be Learned (Carnegie Commission on Preventing Deadly Conflicts, forthcoming, 1997), and From Coup to Coup: Burundi’s Failed Democratic Transition and International Crisis Management (Council on Foreign Relations and Twentieth Century Fund, forthcoming in 1997). Back.

Note 3: See Preventing and Mitigating Violent Conflicts: A Guide for Practitioners (Washington, D.C: Creative Associates International, Inc, 1997) Back.

Note 4: The sort of conflict-sensitive development aid country-level program planning is being tried in modest ways in the Greater Horn of Africa Initiative. Back.

Early Warning and Early Response