By Lincoln P. Bloomfield
[Adapted for the web from an article in Millennium: Journal of International Studies, London School of Economics, 1997, Vol.26, No.3, pp. 709-726.]
Since the Cold War ended, some conflicts have been brought to an end by international mediation and other classic techniques of dispute settlement. Such success stories include, for example, El Salvador, Namibia (former South West Africa), Mozambique, Nicaragua, and Niger. However, if one goes back to the beginning of the 20th century, the great majority of civil wars — 85 percent — ended by military victory.
A compelling reason to keep working at conflict research is to try to change the proportion of wars that proceed as far as military victory, as compared to negotiated settlements. The Computer-Aided System for Analyzing conflicts (CASCON) described below is one of several attempts to contribute to that effort. Realistically, wars usually end when one side wins or the parties experience significant war weariness. Thus, World War I was only a temporary end-of-war because there was no clear-cut victory. The imperfect quality of the armistice helped Adolph Hitler to rise to power with the claim of a 1918 sell-out by those who had led an essentially "undefeated" Germany. On the other hand, World War II was a successful end-of-war because Germany and Japan were decisively defeated militarily, thanks to U.S. President Franklin Roosevelt's controversial insistence on a policy of unconditional surrender. When, in the 1990s, the Palestinian-Israeli, Northern Ireland, and Bosnia conflicts showed halting signs of drawing towards their ends, the key factor was not victory, but exhaustion of the parties and recognition by some of the most zealous partisans that neither side could really win.
Although the second half of the 20th century has been notable for the absence of the kind of "hot" great power wars that bloodied the first half, the Cold War period featured other varieties of conflict that made it both tumultuous and dangerous.
What made the Cold War tumultuous was the decolonization movement that created almost 100 new states. The transition to independence was peaceful in some former colonies such as Ghana, Nigeria, Morocco, Guyana, the United Arab Emirates and Micronesia. However, in others, such as the former Belgian Congo, Cyprus, the Horn of Africa, Algeria and Indochina, it was bloody indeed.
What made the period dangerous was the proliferation of proxy conflicts in countries such as Korea, Cuba, Vietnam, Israel, Angola and Afghanistan. In a situation of mutual nuclear deterrence, these proxy conflicts served as surrogates in the global struggle for power and influence between the superpowers. At the same time, traditional ethnic tensions in the Caucasus, Yugoslavia and Central Africa were contained, held in check by communist rule or by the controlling hand of a superpower. In Russia, Josef Stalin exiled entire groups (such as the Volga Germans) to eliminate that distraction. In Yugoslavia, Marshal Tito pretended that the country was homogenous. And in Africa, French, British and Belgian troops kept the area focussed on the issues of concern to the metropolitan governing authority.
Towards the close of the 20th century there are no more homogenizing empires. In the early 1990s, some local conflict situations that had been suppressed or distorted by the Cold War suddenly rose to the top of the crisis agenda. Some, like Bosnia and Rwanda, began to threaten regional stability. Others, such as Somalia and Haiti, assailed the conscience of a world that watched humanitarian outrages in living color. For a brief moment, the international response to such local conflicts gave shape to what U.S. President George Bush christened a "New World Order."
The notion of intervention by international society for peacekeeping and peacemaking proved premature when, particularly in the United States and the United Nations, a backlash set in against what some called the "Cases from Hell." Peacekeeping developed into "mission creep" towards nation-building and even enforcement. Several scholars and officials who had been deeply involved in those cases reflected on the conditions associated with the termination of hostilities. While anecdotal, in that they focussed on single cases that could not be generalized, these analyses contain much of value and I will return to them.
Those signature conflicts of the 1990s were only the most recent of more than a hundred so-called "small wars" since 1945. Without engaging in the debate over numbers and definitions, a safe round figure for actual wars over the past several decades is plus or minus 100. At a time when most strategic analysis was driven by the over-arching Cold War struggle and the high-profile nuclear strategy of the latter, a substantial amount of scholarly research in the same period was devoted to a more general study of conflicts as a generic phenomenon. Classic attempts at correlation analysis were made by, for example, Quincy Wright, J. David Singer, Ted Robert Gurr, and Rudolf Rummel. Another major research genre focused on conflict resolution, as, for example, the work of Evan Luart, Michael Brecher, Hayward R. Alker Jr., Nazli Choucri, Raimo Väyrynen, Joseph V. Montville, Michael S. Lund and Kuman Rupesinghe.
This paper draws extensively from a policy-oriented research project that I directed at the Massachusetts Institute of Technology (M.I.T.). The project sought to analyze local and regional conflict and, in the process, derived some general propositions on the dynamics of war along with its causes and effects. At the outset, the research staff decided not to include in the database any direct confrontations between the nuclear superpowers, such as the Cuban Missile Crisis, nor what might be termed the "direct proxy conflicts" of Korea and Vietnam, on the grounds that these were idiosyncratic and did not lend themselves to development of broadly generalizable hypotheses that could be applied to new conflict cases.
Other conflicts that were in part Cold War proxies, such as the Arab-Israeli wars, Angola, Laos, Nicaragua, and El Salvador were included because of their intrinsic character that the research staff believed might have entailed violent conflicts even in the absence of the Cold War.
The MIT research on small wars in the 1960s and '70s spun off a computerized conflict analysis system called CASCON. Allen Moulton and I developed the system based on detailed case studies and a dynamic conflict phase model created by Amelia C. Leiss and myself (see Table 1). I recently performed additional research on the database that generated a rank-ordered catalogue of factors associated with ending wars, which is presented here for the first time. After reviewing the CASCON results, I will examine briefly the findings of some other analysts on the termination of wars, and suggest some conclusions for policy. The purpose of this article is to identify factors, other than military victory by one side, that are conducive to terminating hostilities (see definitions below). After analyzing the CASCON database and comparing the results with the findings of other scholars and practitioners, the article concludes that the belligerents need to change their perceptions, and limit their choice of means as well as objectives. Here, third-party governments often have a decisive role to play. Additional outside help in the form of international organizations or ad hoc mediators is also crucial — as is "mutual exhaustion" among the parties — for terminating hostilities.
Before looking at the data, the reader is asked to keep in mind the specific definitions of disputes, conflicts, hostilities (war) and settlement that we believe aid clearer thinking. Phrases such as "war", "conflict", "dispute" and "escalation" are often used interchangeably (for example "trade war", "escalating rhetoric", "conflict between allies and axis in World War II", "bloody disputes" etc.). As with the Red Queen in Alice in Wonderland, these words commonly mean what people say they mean in discussing any and all differences, whether quarrels, political struggles, economic or social differences, coups d'état or world wars. However, they encompass a huge spectrum of the differences between contending parties, starting with simple arguments and ending with wars. One situation might feature bellicose language yet actually be a fairly peaceful scene (e.g. Greece and Macedonia until recently). Another, however, might involve saber- rattling with real sabers (e.g. Greece and Turkey). Finally, something profoundly different takes place when organized units wearing different color outfits start shooting at each other. Thus the need for a common vocabulary.
Political differences of any kind invariably start with a dispute — a quarrel about something. It might be resources, borders, race, tribe, language, religion, minorities, or who is to rule a piece of territory. A conflict begins when one or both or all sides begin to consider settling the dispute with force, that is, one or another party begins to arm seriously, to import offensive weaponry, or to employ symbolic forms of violence. In sum, a conflict begins when the parties begin to sense a change in the atmosphere that makes it qualitatively different than a verbal squabble, but which does not yet involve any organized violence. These categories can be applied equally well to internal as to interstate situations.
When actual fighting breaks out, we term it hostilities. This means more than a political assassination, car-bombing incident, or border skirmish, all of which are frequent features of non-hostilities conflict situations. The term hostilities implies outright, recognizable warfare involving organized sides generating significant casualties, whether the number is 500, 1000 or 1500. The sides may wear different color uniforms as in the Gulf, or they may wear pajamas as in Vietnam or suits as in ex-Yugoslavia. The point is that a shooting war has started. (Having advanced the necessary cautions about loose usage, the rest of our discussion will, for convenience, occasionally use the word "conflict" generically to save time and the patience of the reader.)
The MIT-based research on local conflict mentioned at the beginning of this article is the most systematic attempt with which I am familiar at cataloging and analyzing factors — including war-ending factors — all drawn from the facts of real-life conflicts. First, a brief explanation of the model on which CASCON is based.
The CASCON model, which Leiss and I developed, is based on the premise that conflict is a dynamic process in the sense of passing through some or all of a sequence of distinctive and identifiable stages or "phases." Within each phase can be found a variety of influential elements conveniently described as "factors", such as personalities, relationships, actions, events, perceptions, and other conditions. Some of these can reasonably be assumed to generate pressures moving the situation toward "worsening", that is, of increased violence or its threat; or, conversely, pressures to move the situation in a more benign direction, that is, away from violence. In other words, "factors" combine so as to worsen or improve the conflict and, thus, move the conflict towards or away from "thresholds" between phases in the direction of greater or less violence.
It should be pointed out that avoiding or minimizing violence may not always be the appropriate policy objective if confronted with a greater evil such as genocide or threat of weapons of mass destruction. However, we chose it as a criterion because, in our opinion, it is a morally valid objective in normal circumstances to attempt to reduce violence. Table 1 is a simplified depiction of the so-called Bloomfield-Leiss "dynamic phase model" which indicates the various steps through which conflicts can travel. This represents the framework we employed for our basic conflict research, including the CASCON system. The "factors", symbolized by arrows at the bottom of the picture tend to move the situation in either direction.
Phase 1 is the "dispute" stage, where a divisive issue exists but has not yet been cast by either disputant in terms by which military power becomes significantly relevant. Phase 2 is the "conflict" phase where no serious shooting takes place, but actual warfare no longer seems improbable as a military build-up starts, an arms race develops, or as military forces are deployed with a corresponding intent to use them at some point and as the parties begin to view the situation in potentially military terms. Phase 3 — the "hostilities" phase — is when the disputants have crossed the threshold to actual fighting. Phase 4 — the cessation of hostilities — is an armed truce, so to speak, but with no end to the conflict, let alone a settlement of the underlying dispute. In Phase 5 the dispute remains but the situation is no longer perceived in military terms. Finally, there is the ultimate "Settlement" stage where the underlying dispute, and consequently, the conflict, are settled. The CASCON computerized system contains factors for the first three phases. Factors for phases 4 and 5 have not yet been identified.
The CASCON model of phased conflict also implies a feedback mechanism, not illustrated here, whereby a post-hostilities situation (Phase 4) can loop back to Hostilities (Phase 3), return to Phase 4, and revert to open warfare again, and yet again. This has been the story of the Arab-Israeli wars (which have broken out between Israel and its neighbors in one form or another in 1948, 1956, 1967, 1970, 1973 and 1982), as well as of India and Pakistan (which have gone to war over Kashmir and later Bangladesh in 1947, 1965 and 1971). Cases in Phase 5 can, in theory, also loop back to pre-hostilities (Phase 2) one or more times.
Violence-producing and violence-minimizing factors interact dynamically during phases rather than at a theoretically exact moment of transition (which may not even be possible to identify). Thus, while thresholds are convenient points of demarcation separating phases, the event of transition is itself a product of forces that have been at work throughout the phase.
Some cases of conflict, fortunately, never cross the threshold into an outbreak of hostilities. Others stay for a mercifully brief time in the hostilities phase and then either move through Phase 4, where the issue is pacified but the conflict still exists, or, perhaps, go to Phase 5 where the disputants no longer have any intention of resolving the dispute by military means. In the best of worlds, peace breaks out in the form of Settlement. Some cases may even go directly from the battlefield to Settlement if the results of battle are decisive enough. A depressing number of conflicts linger in Phase 3, moving through sub-phases representing intensification of hostilities (escalation), or perhaps recuperation in a tenuous cease-fire (Phase 4) until, as in the 1948-1982 series of Arab-Israeli wars, hostilities are resumed.
There is, furthermore, no time limit on any single phase. The Cold War was, in effect, a half-century-long Phase 2 conflict between the nuclear superpowers that mercifully remained in the pre-hostilities phase, although, as noted, proxy wars during the Cold War were often bloody. However, in 1947, the Kashmir conflict between India and Pakistan remained in Phase 2 only a matter of days before plunging into open hostilities that have recurred twice and could still return, this time with nuclear weapons in both Indian and Pakistani hands.
CASCON contains a database of 85 post-World War II conflicts (excluding, as explained earlier, direct Soviet or Chinese clashes with the West). Each case is coded in up to three phases for a total of 571 factors. The coding is by three experts (usually two academic or diplomatic authorities on each case in addition to one coding by research staff at M.I.T.). The task for coders is to make the difficult retroactive judgment as to whether the factors that they select as present in "their" case were, in their opinion, influential in "pushing" it across a threshold in the direction of either war or peace. CASCON also contains a brief history, called a "précis", for each database case. Historians commissioned by our M.I.T. project derived the CASCON factor list of 571 generalized propositions from real-life conflict cases. The present 85-case database, which is global in scope and draws on situations from 1945 to the present, is representative of various regions, conflict types and issues in dispute. [See chart of cases by region and type.]
CASCON has two broad purposes. First, to serve as an aid to the memory. CASCON does this by storing, in readily accessible form, a structured inventory of historic events and circumstances that might be relevant to an incipient or exploding conflict situation. We believe that this method can generate a more suggestive list of historically-similar situations than those remembered by distracted crisis managers (or students) in the middle of the night, or bounded by the specialized mental map of a regional expert. It is not that officials are not well informed; on the contrary, they are often extraordinarily so. However, because bureaucrats and diplomats typically pass in and out of their assignments every three years or so, official historic memory is always problematic. CASCON also supplies a convenient way to store information about a new dispute or conflict in a way that can also tell us what one does not know about a case by listing factors coded as "no information", and in so doing, creating a checklist of "intelligence requirements."
Second, CASCON serves as an aid to the imagination. It does this by allowing the user to compare his or her new case at any time with violence-generating or violence-minimizing factors in database cases, supplying clues to suggestive repetitious patterns from recent history.
There is, of course, no reason why CASCON cannot be employed to analyze conflicts in a single region or type. It is a simple matter for the user to limit to a desired sub-set the analysis of cases and the results achieved by the comparison routine. However, one basic premise of CASCON is that, while regional and other characteristics are obviously distinctive, there nevertheless exist some common across-the-board features of conflicts that can help illuminate both the general phenomenon of conflict and the dynamics of a single case.
As an artificial system, CASCON may help jog the memory and prod the imagination, but, of course, it cannot substitute for human knowledge and experience. Moreover, its basic premise — that historical cases drawn from all regions can have common features and in that sense be "similar" — is subject to the caveat that some constellations of events are unique and some local situations are overwhelmingly culture-bound. Having posted this warning label, I would, nevertheless, contend that many aspects of conflict can profitably be generalized across regions and across decades, and that need for better insights into conflict is too important to be imprisoned within the limits of any professional formation or cultural bias.
It has always been our hope that users would experiment with CASCON to develop patterns of factors found to be associated in a significant grouping of cases with, for example, outbreak of hostilities, or dispute settlement, or transition from dispute to conflict, and so forth. My expectation was that such a pattern could constitute an "ideal" or profile case, which could be used as a kind of template against which to match new cases.
In a recent attempt to develop some systematic findings from CASCON, I searched the database to identify factors that have been empirically demonstrated to be favorable to three specific kinds of outcomes. The search discerned factors that, in the "conflict" phase (Phase 2) tended towards violence, and secondly, factors that tended away from violence. In addition, it also discerned factors that, in Phase 3, tended towards the end of the particular wars in question. This involved examining the factors most relevant to the particular outcome and computing their impact on a scale of 1 to 5. The result was a set of three synthetic "profile" cases. The one of relevance here concerned the termination of hostilities.
This Profile case that I designated "PRO3" created an artificial conflict situation in Phase 3 that after relatively brief fighting moved out of the "hostilities" phase. The CASCON database cases that I chose a basis for this exercise were, first of all, three that went to hostilities and then were quickly settled (Cuba 1952-59, Grenada 1983, and Indonesia-Malaysia 1963-65). I selected three additional Phase 3 cases that went relatively promptly from fighting to Settlement, but briefly went through Phase 4 or Phase 5 en route (Algeria 1954-62, Bangladesh 1971, and Congo 1960-63). 66 Phase 3 factors (out of 206) were coded as tending "away" from hostilities in at least 2 of my 6 cases. After weighting the expert codings according to how many cases — out of 6 — they appeared in, and after taking into account whether they were coded "much", "some", or "little" Away, 25 factors scored high enough to be deemed significant. What follows is the rank-ordered list of 25 factors that were most closely associated with termination of hostilities.
Should this list be taken as definitive? Only in the sense that it emerges from a more systematic analysis of history than is customarily available to scholars and diplomats. This list adds up to an "ideal model" and these factors will doubtless never appear together in a single, real-life case. However, if a new conflict case matches even somewhat closely the factor codings of this profile case, it is probably a good bet that favorable conditions exist for terminating hostilities. However, some limitations on the analysis need to be recognized. Above all, it should be remembered that the factors cited are "generic" statements derived from events in a wide variety of actual conflicts studied. Thus, they vary widely in scope and applicability. Some will strike the reader as self-evident. Nevertheless, this set of statements constitutes a checklist of historically validated war-terminating factors that may have potential value to those with responsibility for policy or for analysis. The following is a list of the rank-ordered Phase 3 CASCON factors scoring highest in their association with war termination.
Note that some factors may, on first glance, seem incongruent or counterintuitive — for example, number 6 on new arms may seem to portend a wider war. However, in regards to number 6, coders of specific cases believed that a sudden major arms shipment created the fear of escalation evidently needed to provide impetus for settling the conflict. Similarly, the repressive actions implied in number 16 evidently stimulated compromise, or perhaps even surrender, in a majority of cases chosen. Number 19 suggests that military preponderance can end fighting. Unless one is morally offended by such stark realities, none of these points should be particularly surprising.
For the purpose of comparison, I analyzed a second set of CASCON factors that derived from a review of all 61 CASCON database cases that crossed the threshold into the Phase 3 "hostilities" phase. Unlike the "Profile case", these cases were not necessarily quickly resolved or even resolved at all. The four factors below represent the reply by expert coders to the question "Regardless of the actual outcome of your case, what factors present in the Phase 3 hostilities phase of your case did you judge to have tended away from "worsening", and therefore tending toward peace?"
In this run, no Phase 3 factors were coded as "away" from hostilities for all (or even 75%) of those 61 cases. However, four factors were so coded in, at least, 50% of the 61 hostilities cases. They are rank-ordered as follows:
Number 4, entailing limits on the means and/or objectives of the status quo side was the only one not on the earlier Profile case list of 25 factors. However, it is a prominent feature in other war-ending research summarized in the following section. It was doubtlessly not on the rank-ordered list of 25 factors because it was no relevant to the limited list of six cases used. However, since it is prominent in the list of "other scholars" it clearly belongs on our final catalogue of important war-ending factors.
In order to combine further evidence with the CASCON findings, this section reviews other recent research for factors associated with ending wars. One of the few studies explicitly addressing the issue of why wars end is by Fred Charles Iklé (who later became Director of the U.S. Arms Control and Disarmament Agency and Under Secretary of Defense). Iklé analyzed impulses both to start and end a war. His database comprises primarily the major wars of the twentieth century, as was that of strategic analyst Paul Keckemeti, who noted the major role played by exhaustion in what he called "non-total wars."
Considerably more literature is available on crisis management and the conditions for settling disputes. Some factors had to be derived from the conclusions of authors about dispute settlement and negotiation more generally rather than from answers to the precise question of why wars end. Nevertheless, they help to flesh out the picture suggested by our CASCON histories. Jacob Bercovitch derived findings about mediation from some wars that did end. Some of his general propositions are suggestive. "Mediation must take place at a propitious moment." First there must be some "test of strength" between the parties, then ideally a 12-26 month stalemate or mutual exhaustion. A confirmatory narrative from a long-time British diplomat at the United Nations, Sir Brian Urquhart, described the steps that, in the logic of negotiation, "had" to follow UN Secretary General U Thant's 1965 appeal to India and Pakistan to stop fighting. According to Sir Brian Urquhart, "Each obviously had to let the situation run down a little further and feel more pressure before the war could stop."
John Campbell's study of the Trieste dispute in the early 1950s similarly argued that, in a parallel to the mantra of "location, location, location" by an American real estate agent, the mantra of the conflict-resolver should be "timing, timing, timing." He wrote with regards to Trieste that the atmosphere propitious to a negotiated settlement between Italy and Yugoslavia "did not exist earlier, and ...might not have recurred later." Campbell identified some factors that can apply to negotiating the end of hostilities: a sense of urgency, availability of third parties, involving skilled diplomats with a relatively free hand, secrecy, face-saving provisions, and the so-called Rhodes formula of proximity talks. To these he also added luck.
Two well-known students of history derived relevant lessons about ending wars from a study of great power conflict (World War I, the Chinese intervention in the Korean War of 1950, and the Cuban Missile Crisis of 1963) which also apply to terminating lesser wars. To avoid disaster, one or both sides need to limit their objectives and/or the means employed in their pursuit. In addition they need to maintain top-level control of military options, create pauses in the tempo of military operations, coordinate diplomatic and military moves, keep military moves proportionate in order to avoid giving a contrary impression that can provoke preemption, and choose options that signal a desire to negotiate and that leave opponent a way out.
A more directly relevant source with which to compare the CASCON results is recent commentary by participants on the conditions for ending a given conflict. These are unfortunately few in number, perhaps because most diplomats are not in the habit of publicly criticizing their experience. American diplomats are the exception to the rule of reticence, which may explain why most of the material that follows is by Americans.
One comprehensive study of negotiated terminations of civil wars in the twentieth century by two authors, including the former US Assistant Secretary of State for African Affairs, Chester A. Crocker, concludes that success in "making peace settlements work" is contingent on the following factors:
In successful situations, international bodies played a crucial role in ending military hostilities and defusing tensions. Third parties remained thoroughly engaged through the processes of negotiation and implementation, creating a stable security environment.
In a separate study, Canadian scholar Fen Osler Hampson identified the crucial factors in the most successful recent cases of what he called "peace implementation" — Cambodia, El Salvador, Namibia, and Mozambique. The terms for a political settlement had been worked out in extensive, detailed negotiations. Equally important, all parties agreed to those terms, as did the major regional and global powers.
Important elements of those terms were demobilization, disarmament and restructuring of the armed forces of the parties. Additionally, it also helped that third parties (including international governmental and non-governmental organizations) performed essential monitoring and verification, temporarily taking over basic administrative functions if necessary, and establishing a rule of law through elections, human rights programs and judiciary reform. (Ironically, one of the major elements of the backlash against involvement in the Somalia case was an insistence by contributing states that peacekeepers not get entangled in such "nation-building").
In the late 1980s Crocker analyzed the situation for the southern Africa that were locked in conflict. Crocker, seeing the links between the conflicts in the region, argued that a solution to the long-standing racial conflict in South Africa could benefit from the earlier success in moving Angola and Namibia from war to peace in 1988. Keeping in mind that these are the views of one key individual, the major ingredients of the advice he suggested for the states were: be prepared to yield in order to get what you seek, and act on the principle of a "peace without losers." For Crocker the key factors in bringing about the Namibia-Angola Accords were a realistic framework of negotiation, the availability of a suitable forum, and a basic equilibrium of power that made it possible for every party to gain.
Another experienced American diplomat, Ambassador Charles W. Freeman, elaborated on the Crocker thesis with specific reference to Namibia. By 1987, the infeasibility of a military solution had become clear to both Pretoria and the Southwest Africa People's Organization (SWAPO) movement. As the human and economic costs of war escalated for all parties, they became convinced that there was no alternative to a negotiated settlement, particularly given the linkage to Angola where the nationalists could not achieve their objectives while Cuban and South African forces fought it out. However, the Cubans could not be sent home until Namibia was settled. According to Freeman, the conditions were thus propitious for a deal. The Cubans needed an honorable way out of their quagmire, while Moscow needed to regain its international reputation.
However, before all the tumblers could click into place, the politics of the situation required a final thrust. This took the form of a Soviet-backed offensive against Jonas Savimbi's Uniao Nacional para a Independencia Total de Angola (UNITA) faction in the fall of 1987 which was duly defeated by UNITA with South African help. At that point, according to Freeman, both sides initiated a highly dangerous level of escalation. However, the prospect of escalating the conflict forced them to contemplate a "choice between a negotiated settlement and bloody battles that would not alter [their] basic military balance." Settlement was facilitated by the fact that of the two former superpowers backing the opposing sides, the United States was now available as a mediator, with the positive help of the rapidly-declining Soviet Union. Freemen's analyses, like others in this section, represent the personal judgements of an experience diplomat. Others might have analyzed the same situation differently. The fact remains that the final settlement for both Angola and Namibia embodied the linkage between the two conflicts proposed by Crocker eight years earlier.
The role played by external powers was also crucial in bringing a negotiated end to the sanguinary Sandinista vs. Contra warfare in Nicaragua in the 1980s. Paradoxically, the United States fomented the war to overthrow the Soviet-supported Sandinista regime. The sudden collapse of Soviet external power was one obvious agent in the peace process. Kenneth Roberts identified two other factors that positively affected the conflict. "US policy itself was repeatedly undermined and constrained by the collective action of smaller Latin American countries (in the "Contadora Group") who sought to end the contra war, as well as by the internal contradictions within the U.S. policymaking apparatus." The latter referred to the growing controversy in Washington over Nicaragua, including the so-called "Irangate" scandal of illegal diversion of funds to the Contras.
Specific factors identified by Kenneth Roberts included unprecedented Latin American regional foreign policy cooperation driven by fear that the U.S. would directly intervene and regionalize the Nicaraguan war. In addition, by 1987-88, according to Kenneth Roberts, "the possibility of a military solution [was] increasingly remote."
If we take together the quite disparate analyses of individual cases reviewed by these analysts, we can nevertheless come up with a rank-ordered list of the factors they identify, and compare them with our CASCON-generated factors. Their eight highest-scoring propositions follow:
One can readily see that these recurrent factors identified by other scholars and practitioners converge with our CASCON ranked factor list on four key propositions:
To these I would add, from research cited at the opening of this article, the element of mutual exhaustion of the parties — a factor that seems obvious on the basis of common sense.
In much of the CASCON historical database, wars frequently end as a result of military victory rather than negotiation. However, students of conflict in the present era are likely to consider a war-ending strategy of decisive victory to be morally repugnant. This is evidence by the fact that we rarely find that factor among those they consider as solutions to war. Did our CASCON historians look at history with a more cold-blooded eye because the events in question were safely in the past, and so the historian could be morally neutral? Are contemporary "peace researchers" likely to distort their analyses because, so to speak, Carthaga delenda est is not the message from the field that peacemakers and negotiators want to hear? The scholar must go where the evidence leads. However, I, for one, strongly share their moral preference for negotiated settlements.
[My greatest debt in the research drawn on here is to Amelia C. Leiss and Allen Moulton.]
[Lincoln Palmer Bloomfield is Professor of Political Science Emeritus at the Massachusetts Institute of Technology. He served in the U.S. State Department as policy planner on UN affairs, and in the National Security Council as director of global issues. Author of a dozen books and over a hundred articles, his most recent book is Managing International Conflict: From Theory to Policy in collaboration with Allen Moulton, published by St. Martin's Press, and including the award-winning CASCON software program.]
 Roy Licklider, "The Consequences of Negotiated Settlements in Civil Wars, 1945-93", American Political Science Review (Vol.89, No.3, 1985), p.686. In his article Licklider credits the work of Prof. Stephen John Stedman of Stanford University for some of his data.
 See Lincoln P. Bloomfield and Allen Moulton, Managing International Conflict: From Theory to Policy (New York: St. Martin's Press, 1997).
 See for instance the author's "The Premature Burial of Global Law and Order: Looking Beyond The Three Cases From Hell", The Washington Quarterly, Summer, 1994.
 Many analysts define "wars" as entailing at least 1000 deaths. Using that criterion means 41 recent wars plus 70 minor ones. Some using the same criteria counted around 90 — 53 "major conflicts" in countries plus another 37 involving political violence. Of the total, over 80 percent were in the developing regions, only a handful were inter-state and the rest were internal. Other analysts counted 43 "small wars" in 1993. Still others use larger numbers which turn out to include skirmishes, bloody coups d’état and similar "sub-limited" unpleasantnesses.
 Quincy Wright, A Study of War, Revised Edition (Chicago, IL: University of Chicago Press, 1965); J. David Singer, The Correlates of War (New York, NY: Free Press, 1980); Ted Robert Gurr, Minorities at Risk (Washington DC: United States Institute of Peace Press, 1993); and Rudolph Rummel, Dimensionality of Nations Project (Ann Arbor, MI: Inter-University Consortium for Political and Social Research, 1976).
 C. Evan Luard, Conflict and Peace in the Modern International System (Albany, NY: State University of New York Press, 1988); Michael Brecher, Crisis, Conflict and Instability (New York, NY: Oxford University Press, 1989); Hayward R. Alker, Jr., International Conflict Episodes (Ann Arbor, MI: Inter-University Consortium for Political and Social Research, 1980); Nazli Choucri and Robert North, Nations in Conflict (San Francisco, CA: W.H. Freeman, 1975); Raimo Väyrynen, "Domestic Crises and International Wars" in Peter Wallenstein (ed.), Peace Research: Achievements and Challenges (Boulder, CO: Westview Press, 1988), 70-102; Joseph V. Montville, Conflict and Peacemaking in Multiethnic Societies (Lexington, MA: Lexington Books, 1991); Michael S. Lund, Preventing Violent Conflicts (Washington DC: United States Institute for Peace Press, 1996); and Kuman Rupesinghe, Toward a Policy Framework for Advancing Preventive Diplomacy (London: International Alert, 1994).
 On the 1960s and 1970s MIT research, see Lincoln P. Bloomfield and Amelia C. Leiss, Controlling Small Wars (N.Y.:Alfred Knopf, 1969). On CASCON, see Bloomfield and Moulton, op.cit., in note 2.
 See also Albert Legault, "United Nations Peacekeeping and Peacemaking" in The State of the United Nations, 1992 (Providence: Academic Council on the UN System), Reports and Papers No.3, 1992, p.30.
 The following section is based on pages 99-103 in Bloomfield and Moulton op. cit.
 See Bloomfield and Leiss, op.cit. in note 7, pp. 45-50.
 Our use of "status quo" and "non-status-quo" sides refers not to ideology but to which group possessed the power/resources/territory, and which group sought to overturn that status.
 Fred Charles Ikle, Every War Must End, (New York: Columbia University Press, 1971) p. 34.
 Paul Keckemeti, Strategic Surrender, (Stanford: Stanford Univ. Press, 1958), p. 17.
 Jacob Berkovitch, "International Mediation: A Study of the Incidence, Strategies and Conditions of Successful Outcomes," Cooperation and Conflict, XXI, 1986, p. 161.
 Brian Urquhart, A Life in Peace and War, (New York: Harper & Row, 1987), p. 206.
 John Campbell, Successful Negotiation: Trieste 1954 (Princeton: Princeton University Press, 1976), pp. 146-153.
 In negotiations in 1948-49 between Arabs and Israelis on the island of Rhodes, relations between the parties were so bitter that the UN Mediator had to shuttle between separate hotel rooms (thus "proximity talks") rather than negotiating around a common table.
 Campbell, op.cit., in note 17 pp.146-53.
 Gordon A. Craig and Alexander L. George, Force and Statecraft (New York: Oxford Univ. Press, 1983), pp. 206-07.
 Chester A. Crocker and Fen Osler Hampson, "Making Peace Settlements Work", Foreign Policy (No. 104, 1996), pp. 70-71.
 Fen Osler Hampson, Nurturing Peace: Why Peace Settlements Succeed or Fail (Washington DC: US Institute of Peace, 1996).
 Chester A. Crocker, "Southern Africa: Eight Years Later", Foreign Affairs, (Vol.68, No.4, 1989, pp. 144-64.
 Charles W. Freeman, Jr., "The Angola-Namibia Accords", Foreign Affairs (Vol.68, No. 3, 1989), pp. 126-41.
 Ibid., p. 134.
 Chester A. Crocker, "South Africa: Strategy for Change", Foreign Affairs (Vol. 59, No. 2, 1980-81).
 Kenneth Roberts, "Bullying and Bargaining: The United States, Nicaragua, and Conflict Resolution in Central America", International Security (Vol. 15, No.2, 1990), p.100.
 Ibid., pp.91, 93.
 Ruth Sivard in her annual accounting of World Military and Social Expenditures defines war as entailing "deaths averaging more than 1,000 per year." Her lists cover most of the conflicts we and others have studied. (Leesburg, VA: WMSE Publications, annually).
 Karel Lindgren, G. Kenneth Wilson and Peter Wallensteen, "Armed Conflicts over Government and Territory", in States in Armed Conflict 1988, ed. Peter Wallensteen (Uppsala University, Department of Peace and Conflict Research) Report No. 30, July 1989, p.35.
 Human Development Report, UN Development Program (New York: United Nations, June 1994), p. 47. The Report adds that 90% of the casualties suffered were civilian.
 SIPRI Yearbook 1994 (New York: Oxford University Press, 1994).
Copyright © 1997-99 Lincoln P. Bloomfield and Allen Moulton