2
Armed Conflicts and Peace Agreements
2.1 The Concept of Conﬂict
Astrong statement is that conﬂicts are solvable. This is not necessarily an
idealistic or optimistic position. As this book will show, it is a realistic
proposition. Most actors in conﬂicts will ﬁnd themselves in need of negotiations
at one time or another. Even if a conﬂict results in war and destruction,
there may have been other options and alternative paths for the conﬂict. There
are frequent statements on the inevitability of conﬂict, violence and war.
Indeed, ﬁnding solutions may often be difﬁcult. This difﬁculty not only arises
out of political constraints, but can also be due to a lack of insight or imagination.
There are also views of the desirability and even necessity of violence and
war. Unbearable conditions or impossible threats may make such opinions
understandable. Too often, however, the results of war negate the very hope
for a better future that may initially have motivated the war. Few wars follow
the paths anticipated by the actors. Short wars may avoid such pitfalls, but
who is to guarantee that a war will be short? Many wars have started from
this premise. Afterwards, it will be asked: were all avenues used to ﬁnd a
peaceful solution prior to the initiation of war? Only after this can be convincingly
proven do the arguments of inevitability and desirability approach
validity. Thus, the determined search for a solution is not only a moral question;
it is also a rational one. This is the sole way in which a free society will
be prepared to accept the strains of war. Indeed, if conﬂicts are exposed to
such early challenges, solutions may actually be found, even in unexpected
situations. Thus, conﬂicts are solvable and there are many and varied experiences
of such solutions.
If conﬂicts are solvable is it also true that conﬂicts – sooner or later – will be
solved? Clearly, once a conﬂict has developed into a war the options are fewer.
At that moment, the primary actors will pursue victory rather than a joint
solution. The victory of one side over the other is, then, a possible outcome,
even to the point of the other’s capitulation, dissolution and disappearance as
an actor. The record shows that this is what happens in some conﬂicts, but by
no means all. Conﬂicts will come to an end at some point in time. Whether
that ending is a solution, a victory or a stalemate has to be scrutinized. To this
should be added the question of whether the conﬂict is likely again to be
armed and violent.
Victory is the outcome preferred by most actors. If achieved, it may solve
parts of an issue, but often not the entire problem at hand. The victory of the
allied countries over Nazi Germany is a case in point. After the failure of
the agreement reached in Munich in 1938, it was no longer possible for the
Western powers to consider negotiations with Adolf Hitler and the Nazi
regime. The end of the Second World War meant the implementation of the
demand for unconditional surrender and the elimination of the Nazis as an
actor. This was as clear-cut a victory as can be. It did not, however, mean the
end of Germany. The issue of Germany’s position in the international system
still had to be settled. Conﬂict among the victors arose over this question. It
became one of the few core issues in the Cold War. A solution developed as
new leaders emerged in West Germany. They were democratically inclined,
conscientiously building on pre-1933 democratic traditions and new ideas
from the Western powers. A reintegration of Germany into the international
system took place, ultimately even allowing for its reuniﬁcation in 1990, but
only 45 years after the end of the Second World War. It was possible with a
new Germany, willing to admit its responsibility for the past and able to
accommodate to the present. If the Second World War had been a question
solely of Germany’s role in the international system, there was a route through
peaceful dialogue and development. A solution within a democratic framework
among democratic countries was found. It could have been found before
the Nazis took power. For any country, there are, in other words, always
alternatives to a war strategy for achieving goals. Regimes, however, may
deliberately narrow those choices and construct situations where the outcome
becomes only one: defeat or victory. Nazi Germany chose such a path.
Conﬂict precedes conﬂict resolution. There is already considerable analysis
of the origin and the pursuit of conﬂict. Machiavelli and Clausewitz are
important writers in one Western tradition of conﬂict analysis. Adam Smith
and Karl Marx offered competition and class analysis as other tools for
understanding. In classical Chinese discourse Sun Tzu is a central writer, as is
Kautilya in India. Military-strategic thinking has become universally shared,
and there is often, among military ofﬁcers, a surprising degree of common
understanding across battle lines. Also, the analysis of societal contradiction
has such cross-cultural traits, Smith and Marx being inﬂuential in different
quarters across the globe. The same, however, is not true for conﬂict resolution
thinking. It is a novel topic. It is less developed and less coherent. Thus, it
is important to introduce the ideas of modern thinkers. It is also necessary to
relate them to trends of social science thinking.
14 UNDERSTANDING CONFLICT RESOLUTION
‘Conﬂict’ has many meanings in everyday life. To some it refers to behaviour
or action. There is conﬂict when a trade union goes on strike or an employer
locks out its employees. It is also conﬂict when two states are at war with one
another, and where battleﬁeld events determine their relations. The actions
constitute the conﬂict. If this were all, however, it would mean that a conﬂict
ends once this behaviour ends. Few would agree to this. A cease-ﬁre is not the
end of a conﬂict. Even the cession of verbal statements, non-violent actions,
the mobilization of petitions, demonstrations, boycotts and sanctions may
only indicate that there is an interlude in the conﬂict. Actions may resume at
some later stage. There may still be dissatisfaction. Obviously, conﬂict is more
than the behaviour of the parties.
A closer look indicates that the parties in an industrial dispute will not cease
their actions until there is some movement on the issues which sparked the
dispute. The ‘issue’ refers to the incompatible positions taken by the parties,
motivating their actions. This, then, is a deeper understanding of what a
conﬂict is. It contains a severe disagreement between at least two sides, where
their demands cannot be met by the same resources at the same time. This is
an incompatibility. Positions are incompatible. There is some form of scarcity.
If there is an abundance of resources, the demands from the various sides may
easily be met. The incompatibility can be solved. If there are limited resources,
however, problems will arise. The easy solutions are no longer available and
more ingenious ways have to be found. How this can be done will be discussed
later. For the time being it is sufﬁcient to note that when the parties
adjust their demands so that there is no longer scarcity, the conﬂict disappears.
The incompatible demands have been handled. Incompatibility appears to be
a key to the existence of conﬂict. If there are no actions although it is possible
for an outsider to point to incompatibilities, there is a latent conﬂict. Manifest
conﬂict requires both action and incompatibility.
This is still not enough to get an initial understanding of the concept of
conﬂict. We need to include the actors as well. Many would say that trade
unions are created to make conﬂict. This is why they have a membership.
Members expect to be protected even to the point where a manifest conﬂict
becomes a distinct possibility. This means that there is a tension built into
the relationship between the employer and the employees. ‘Conﬂict’ does
exist, even if there are no actions taken or demands formulated. The conﬂict is
internal to the system. Similar descriptions also can be found for the interstate
system. It is argued that sovereign states are inevitably locked into conﬂict
with one another. States are continuously preparing to defend themselves from
possible attack in order to protect their own survival. Such preparations only
conﬁrm to others that there are real dangers, thus they do the same. These are
the dynamics of the well-known security dilemma (Herz 1950; Waltz 1959,
1979). This perspective suggests that the existence of one state is a danger to
any other state. As long as there is unpredictability in the system, there will be
Armed Conﬂicts and Peace Agreements 15
fear and, thus, conﬂict. For our purposes it means that actors or parties are
fundamental for conﬂict to exist. If the actors are formed, and if they make an
analysis where their needs for survival are in conﬂict with others, then there is
conﬂict built into the system. This means that the history of the actors, the
actors’ understanding of their own role and their resources are important
elements in conﬂict analysis.
From this we can conclude that conﬂict consists of three components:
action, incompatibility and actors. Combining them, we arrive at a complete
deﬁnition of a conﬂict as a social situation in which a minimum of two actors
(parties) strive to acquire at the same moment in time an available set of
scarce resources. This deﬁnition brings together essential elements from a
number of commonly used deﬁnitions. It includes the actors or ‘parties’ in the
deﬁnition which, as we have just seen, is basic. In many deﬁnitions the actor is
left as a separate item. However, the preceding arguments make clear that it is
integral to the analysis and to the deﬁnition.
The word ‘strive’ in the deﬁning sentence requires a comment. It is a vague
term, but the point is that when the parties are acting, they are doing something
(however minimal) to acquire the resources. ‘Strive’ may even include
warfare. It covers a wide range of activities.
An additional phrase needs a comment. It is said that the parties are striving
to acquire the resources ‘at the same moment in time’. This is sometimes
overlooked in deﬁnitions and may, again, be self-evident. If one actor is
satisﬁed with having its demands met a year from now, other actors may be
able to meet their goals today. There is no conﬂict today. Perhaps the ﬁrst
actor will worry for the future – will there be anything left? – but if the party
feels it has guarantees, the incompatibility is gone. It is clearly a different
matter when the demands are geared to the same moment in time. It is
conventional wisdom that only one person can be prime minister at a time and
that only one country can have formal jurisdiction over a particular piece of
territory at a time. These resources are regarded as indivisible, for the time
being. If this is what the parties believe, then this is their reality. In actual life
there are solutions even to such problems, for instance, the creation of posts as
ﬁrst and second prime minister (as in Cambodia in the 1990s) or ﬁnding forms
of shared rule for a territory. Such solutions emerge only if the parties perceive
an incompatibility to be divisible. Time, as we notice every day, is scarce but
still has this quality of divisibility, something that our calendars make clear.
Schedules may dissolve incompatibilities.
The notion of an ‘available set of scarce resources’ should not be taken
to include only economic matters. The term ‘resources’ covers all kinds of
positions that are of interest to an actor. To be a prime minister, to control a
particular piece of territory, to be able to propagate a particular idea in the
media can all be covered by the notion of ‘available resources’. This deﬁnition
demands that something is desired which is scarce, be it positions of power,
16 UNDERSTANDING CONFLICT RESOLUTION
attractive land, or access to airwaves. Such resources can sometimes be
estimated in money, square metres or other numbers, but often they are
intangible. For instance, demands for recognition, acceptance of responsibility
for destructive actions or psychological retribution exemplify intangible
values. They are still highly important. They may involve admissions that have
implications for an actor’s standing nationally or internationally, but only
indirectly relate to material resources. Thus, there are incompatibilities relating
to matters of justice, moral norms and guilt.
With the conﬂict concept hopefully clariﬁed, we move to the most difﬁcult
of all conﬂicts: wars. They are different from all other conﬂicts in that they
are irreversible actions. Wars involve the taking of territory, the eviction of
inhabitants, the death of soldiers and civilians, the destruction of property,
resources and the environment, and the disruption of people’s mental, physical,
economic and cultural development. War is among the most destructive
phenomena that one human group can inﬂict on another. In the same category
of extreme conﬂict we also can locate systematic repression, totalitarianism
and genocide. These are actions initiated by human beings. These are matters
that can be ended and remedied by humans, but not undone. They become
strong and conscious elements in the history of peoples, groups and individuals.
Let us ﬁrst look more closely at the exact meaning of war and then
proceed to study recent trends in armed conﬂict and war.
2.2 Identifying Armed Conﬂict
Three projects
A commonly asked question is whether conﬂict and war have become more
frequent and are more destructive today than they used to be. It is a question
about quantity, where it is assumed that conﬂicts are easily comparable. The
question is asked to reach an understanding of where the world is headed, as
a whole, for a particular region or for a particular phenomenon (for instance,
arms production). It is often a question about the future, not only about
history. At the same time, there are those who resent having ‘their’ conﬂict
compared to other situations. Each conﬂict is unique and has its own characteristics.
There are qualities which make them different. The question of frequency
makes little sense to those who are parties in conﬂict. Why should they worry,
it is bad enough with one conﬂict, they would say.
Both perspectives are valid. The projects that exist within the peace research
community all aim at understanding why conﬂicts occur or how they can be
terminated. Their answers to the questions of frequency of wars are actually
by-products of other ambitions. The questions are nevertheless important and
contribute to the development of deeper answers. If there are general patterns
Armed Conﬂicts and Peace Agreements 17
recurring over a large number of different conﬂicts, it suggests something that
can possibly explain why wars begin. By implication it may yield ideas for
improving the situation. Certain factors can be singled out for closer analysis.
Questions of frequency interest the media and the public for other reasons.
Today it is frequently asked if there is a difference between the post-Cold War
years and the Cold War period. Changes in the international system or in
domestic policies, associated with the ending of this confrontation, may explain
our present predicament. There are many other candidates for possible
causation, however, and the impact of each may be difﬁcult to disentangle. In
the analysis many factors are mentioned, such as bipolarity, deterrent effect of
nuclear weapons, changed roles of international organizations, democratization,
the spread of free market mechanisms, changes in media access, concern
for human rights, growth of non-governmental organizations, etc. Comparisons
across time can illustrate a number of effects. They do not necessarily
prove them, however. To be scientiﬁc evidence, cases and periods have to be
selected with rigour and there have to be many observations. For the purpose
of this book it is important to have a general idea of the frequency and
severity of armed conﬂicts in the world. It helps to set the topics of conﬂict
resolution in perspective. Thus, let us review some ongoing efforts.
Armed conﬂict patterns are mapped continuously by several projects. This
book uses the work of the Uppsala Conﬂict Data Project, based at the Department
of Peace and Conﬂict Research, Uppsala University, in Sweden. Data
are published by the department in the annual publication States in Armed
Conﬂict. It is also available through the annual publications of SIPRI, The
Stockholm International Peace Research Institute (for major armed conﬂicts)
and the Journal of Peace Research (JPR), from the International Peace
Research Institute, Oslo, PRIO (all armed conﬂicts).
Second, there is the project on wars and armed conﬂict by the Causes of
War project at the University of Hamburg (AKUF, from its name in German),
mapping the global record of local wars since 1945. This is also published in
annual reports and books (Gantzel and Schwinghammer 2000). The third
enduring project is the Correlates of War project (known as COW), University
of Michigan, Ann Arbor, Michigan, carrying information on wars since 1816.
This information is normally available through databases, sometimes also in
printed publications. It has found a wide use in research projects and the
ﬁndings have been systematized (Geller and Singer 1998; Vasquez 2000).
There are additional important projects that aim to highlight the present
dilemmas of war and violence. Among these is PIOOM in Leiden, the Netherlands,
which also includes human rights violations and produces global
conﬂict maps. There is the Minorities at Risk project, University of Maryland,
focusing on a subset of conﬂict: those involving ethnic minorities around the
world. It contains data for 275 minority groups which have been involved in
some form of conﬂict since 1945. It also includes systematic data on all
18 UNDERSTANDING CONFLICT RESOLUTION
conﬂicts since the Second World War (Gurr 1993, 2000a; Gurr et al. 2001;
Marshall 1999). Crisis behaviour between states can also be used to discuss
questions of frequency (Brecher 1993). A number of researchers have their
own systematic collections of conﬂict-related information which are reported
in international journals (Bercovitch 1996; Carment 1993; Carment and
James 1995; Gibler 1999a, 1999b; Goldstein 1992; Holsti 1991; Levy 1983;
Licklider 1995; Luard 1986; Morton and Starr 2001; Tillema 1989).
The ﬁrst three projects stand out, however, as the most consistent speciﬁcally
addressing armed conﬂicts of all categories, whether between or within
states. They contain additional information which is useful for theorizing on
the war phenomenon. Two are oriented towards understanding the causes of
conﬂict, as can be surmised from the names of the projects (the Hamburg and
Michigan projects). One deals speciﬁcally with conﬂict resolution (the Uppsala
project). What, for instance, do they tell us about trends in armed conﬂict and
war?
The question is simple but requires an understanding of key deﬁnitions
before an answer can be given. The deﬁnitions of conﬂict and war guide the
types of information any project will collect. Conﬂict data projects can potentially
show different global tendencies, depending on what categories of conﬂict
they focus on. The comparability, in other words, can be limited. Furthermore,
there are several criteria that have to be met for a conﬂict data set to be
reliable. First, it must have a deﬁnition that is general, and goes beyond what
is important only to a particular period in history. The deﬁnitions in these
projects meet these criteria: they do not vary with time or with the phenomena
studied. The projects may still be relevant for other concerns as well. For
instance, although ethnic conﬂict is not used as a category in these three
projects, it is possible to retrieve data from them which are relevant for the
study of ethnic conﬂict. There are separate categories of internal war, for
instance. In this way, the projects cover a wider range of conﬂict than does, for
instance, the Minorities at Risk project.
Second, there has to be a deﬁnition that captures conﬂict between as well as
within states. It means that it has to tap the general issue of violence, cutting
across particular legal categories. This allows for an understanding of war
beyond the category of interstate events. Clearly, data on interstate conﬂicts
are more easily compiled. What two states do to one another that might lead
to war is of interest to the surrounding community as well. Thus, such
disputes will have more attention. Conﬂicts inside a state, however, are not as
likely to affect neighbours, thus threatening to make the international recording
of such conﬂicts more sporadic. A full study might require intimate
knowledge of all countries in the world. Thus, it still has to be the ambition to
include all conﬂicts. This is a third criterion which is necessary if changes in
armed conﬂict over time are to be meaningfully discussed. Fourth, the deﬁnitions
have to be precise, so as to guide data collection (operationalization) and
Armed Conﬂicts and Peace Agreements 19
delimit a particular conﬂict in time and space (beginning, ending) from other
conﬂicts. Finally, the data must be open to scrutiny by other researchers.
The Michigan and Hamburg projects
The Michigan project is the oldest and serves as a reference point for many
projects. It was initiated in the middle of the 1960s by J. David Singer and
Melvin Small and is still maintained, something which is an achievement in
itself. It contains data on wars since 1816 and its record is constantly updated.
Basic to COW is the delineation of an international system consisting of
states. Thus, wars are conﬂicts between states where at least two are members
of the international system. In addition, there are extra-systemic conﬂicts,
where only one state is a member. The military hostilities that are included are
those which have led to at least 1,000 combat fatalities during the course of
the conﬂict. The ambition was to have a deﬁnition that captures all signiﬁcant
interstate wars. The civil war deﬁnition was developed later and is comparable,
but not identical. It has a minimum of two parties where one is a
government, but the 1,000 battle fatalities criterion is calculated per year and
does include civilian casualties. It is more likely that, on the one hand, smaller
interstate wars are entering into the statistics than civil wars. On the other
hand, the categories of fatalities may mean that more casualties are reported
for civil wars. The net effect of these differences is not simple to determine.
From 1816 to 1992 it results in a total of 75 interstate conﬂicts, 134 extrasystemic
wars and 151 civil wars (COW WebPages, 25 January 2001). It
identiﬁes a total of 360 wars for the period, an average of slightly more than
two wars or civil wars starting somewhere in the world every year. War, in
other words, is shown to be a pervasive and global phenomenon.
The Hamburg project was started by Istvan Kende in Budapest. It was later
modiﬁed and developed at the University of Hamburg, through the efforts of
Klaus J¨urgen Gantzel. Its results are different. Its deﬁnition does not require
that a party be a member in the international system for a conﬂict to be
included. The AKUF project has, however, the criterion that a state should be
the actor on one side in a conﬂict. The actors should have, at a minimum,
central command and practical control over the ﬁghting. It is also stipulated
that there has to be a measure of continuity in battle. There is no requirement
for a particular number of deaths, which is an important consideration in the
Michigan project (and in the Uppsala project as well, as we shall see). The
Hamburg project, in fact, regards this criterion as a questionable indicator for
practical, theoretical and ethical reasons. For instance, it is argued that
information on deaths is unreliable and that there is no reason to include only
those who have died from battle, but not those who have suffered from other
consequences of the war (Gantzel and Meyer-Stamer 1986: 4–5; Jung et al.
1996: 52). Instead, the ‘continuity’ in the struggle is decisive for inclusion of a
20 UNDERSTANDING CONFLICT RESOLUTION
particular conﬂict. This criterion, of course, results in a problem of judging
continuity in a reliable way.
These criteria mean that AKUF covers a broader set of cases than COW.
The project has data on all wars since 1945. For the period to 1997 the
project reported 197 wars (AKUF WebPages, 25 January 2001). On average
this gives almost four new wars per year. The effect of the deﬁnitions can be
seen more clearly by comparing the years where the projects overlap. For an
almost identical period, that is 1945–92, COW reports 23 interstate wars, 24
extra-systemic conﬂicts and 80 civil wars, a total of 127 wars. It gives an
average of close to three new wars per year. The two projects clearly overlap,
but still AKUF reports more activity. It could mean that the projects do not
include the same major conﬂicts, although this should not have such a strong
impact, as the number is limited. More likely is that many armed conﬂicts are
below the threshold of 1,000 battle-related deaths. Thus, a considerable
number of conﬂicts are not covered in the COW project, although the difference
might have been expected to be even larger.
In its studies of a separate category of conﬂicts, militarized interstate
disputes, COW has accumulated information which corrects for this effect.
This category, which is also of great theoretical signiﬁcance, covers relations
between states. It includes more conﬁned events, such as military interventions,
limited wars and threats of war. Together with the war data, this
gives a more comprehensive picture for relations between states. If these data
are added, the difference between COW and AKUF might be reduced. There is
no record, however, in the COW project of militarized disputes within states.
For both projects internal or civil conﬂicts take up a large percentage of all
events recorded. To develop a deﬁnition that parallels militarized interstate
disputes for intrastate conditions is a cumbersome task. A very large number
of episodes would have to be scrutinized for possible inclusion. It would,
needless to say, be difﬁcult to make a global comparison, as unbiased information
is harder to obtain the more limited the episodes are. For instance, threats
to use force in internal affairs may involve military as well as police forces.
Such threats can also be issued by opposition groups with limited credibility
and representativity. Thus, drawing the lines of inclusion will require additional
distinctions. It is, however, possible to do, for instance, by relying on
data on human rights violations or other indicators of repression.
Both AKUF and COW are oriented to searching for the origins of violent
conﬂict. The difference in approach is partly a reﬂection of distinct theoretical
concerns. COW focuses on understanding interstate conﬂict, and particularly
aims at questioning or modifying so-called realist thinking. This means it is
designed to understand factors such as balance of power, military capabilities,
interdependence and other variables of importance for the working of the
international system. There is, deliberately, no coherent theoretical perspective
guiding the project. Instead, there is a conscious methodological approach.
Armed Conﬂicts and Peace Agreements 21
Reality, as expressed in the data, will speak for itself. It shows how the world
actually functions: correlations are important, thus the name of the project.
This is an empirical approach, where theory development will build on what
has been proven to be repeated and veriﬁed ways in which states really
behave. Theoretical assumptions that are common in realist thinking are
tested against observable patterns of conduct. An advantage with this open
position is that it also makes COW data useful for other purposes. The concepts
and their operationalization are explicit and simple, constructed to
reﬂect world developments over close to two hundred years. COW’s information
has been used for very different investigations. For instance, there is
research on whether arms races lead to war (this can be studied by using the
militarized disputes and comparing them to the war data, with the original
work done by Michael Wallace in 1979). The data are also used for analysis of
the hypothesis on peace among democratic states, resulting in the muchdebated
democratic peace proposition (a large number of articles is devoted to
this puzzle, the early phases of which were crystallized in work by Bruce
Russett in 1993).
The Hamburg project departs from a fairly coherent theoretical approach.
It relates the onset of war to the development of capitalist societies, and sees
conﬂict as a result of the new forms of production, monetarization of the
economy and the resulting dissolution of traditional forms of social integration.
The large number of conﬂicts in the Third World ﬁts with this relationship.
As the project reports that there is an increasing frequency of conﬂicts
since 1945, researchers also conclude that ‘the contradictions in world society
are increasing’. In an interesting twist, relevant here, the authors point out
that even a phenomenon such as ‘ethnic’ conﬂict is a result of processes in
‘which all social mechanisms that previously allowed us to live together are
destroyed’ (Jung et al. 1996: 52–61). In other words, conﬂict resolution
instruments are being eliminated and this makes armed conﬂict more frequent.
The issue of such social breakdown has lately captured considerable
attention, with the term ‘state failure’ as a central concept (Esty, et al. 1998;
Zartman 1995b, 2000).
This perspective points to the difference between the two projects. The
Correlates of War project departs from the notion of a system that consists of
a larger number of independent states. It is a system that does not have central
institutions and lacks means to maintain or enforce decisions for all. Thus, it
is a picture of an uncoordinated world that is the point of departure. In this
world states maintain some predictability through their actions. When war
breaks out, it is not the result of an international society breaking down, as
there is no assumption about such a society in the ﬁrst place. The problem is
found in the strategies for survival used by different states. War is a result of
failed choices, but also the conditions, which make it difﬁcult for states to
pursue other options. In some writings inspired by the project, the importance
22 UNDERSTANDING CONFLICT RESOLUTION
of norms comes forth as an important conclusion, based on the experiences
that peace nevertheless exists in the system (Vasquez 2000). An implication
may eventually be the need for constructing an organized international order.
The Hamburg project, on the contrary, departs from an understanding of
an international system that is fairly integrated, almost having a purpose of its
own, which is to promote market economy and democracy. It is a highly
hierarchical world, centring on the strongest actors in the system, the Western
countries. These actors are also inﬂuenced by the strength of the system. They
are all capitalist, market-oriented and expansive, furthering a system of
asymmetrical linkages. This international system penetrates into all parts of
the world, creating instability and pushing aside traditional forms of social
relations. There is an asymmetry between stronger actors that beneﬁt from
this development, and weaker actors that risk becoming marginalized. The
project has a critical attitude to the basis of the existing international system.
In this way, the disagreement on whether fatalities should be used as a
criterion has a deeper meaning. The Hamburg project could argue as follows:
if one side is vastly superior, it can win an armed conﬂict within a short period
of time, and thus the casualties will be limited. It is still a military operation
for purposes that might be the same as those found in more protracted and
devastating conﬂicts. From the point of view of causes of war, in other words,
the magnitude may not be so signiﬁcant. For the Michigan project, with its
elaborate levels of battle-related deaths per month, for instance, very large
confrontations are the most interesting. The destruction in itself makes them
more important. They suggest inadequacies in balance of power thinking and
deterrence strategies. Such realist theories are developed exactly to prevent
major disasters. If these still occur, the project can show this, and take a
critical attitude to this particular aspect of the international system and its
interpreters. The two projects contrast on important points of departure, their
epistemology. This affects their deﬁnitions for data collection and interpretation
of the resulting data. Thus, both projects are needed and valuable.
Together they highlight different sides of the contemporary global system.
2.3 Trends in Armed Conﬂicts
The Uppsala Conﬂict Data Project
The Uppsala project uses the concept of ‘armed conﬂict’ and focuses on
conﬂict resolution. Like AKUF, it reports annually on ongoing conﬂicts and
has much current information. Its deﬁnitions and understandings of conﬂict
put it somewhere between the Michigan and Hamburg projects. It treats, as
does the Hamburg project, all conﬂicts in an identical way, whether they take
place between or within states. The same deﬁnition applies to both situations.
Armed Conﬂicts and Peace Agreements 23
It means that the distinction between an international system and an intrastate
system is not of primary importance. What counts is the use of violence. The
conﬂicts included are those that have at least one state or government as a
party. This is also the case for the other two projects. It covers conﬂicts from
a threshold level of 25 battle-related deaths in a conﬂict in a year. This is an
easily identiﬁable criterion and requires less evaluation by the researchers than,
for instance, the continuity criterion used in the Hamburg project. It means also
that there is a way of discussing intensity in conﬂict, as is done in the Michigan
project. There are two thresholds (25 and 1,000 battle-related deaths, respectively),
resulting in three categories of intensity: minor armed conﬂicts, conﬂicts
with more than 25 deaths, but less than 1,000 for the year and for the duration
of the conﬂict; intermediate armed conﬂicts, conﬂicts with more than 25 deaths,
less than 1,000 for a year, but more than 1,000 for the duration of the conﬂict;
and wars, conﬂicts with more than 1,000 battle-related deaths in one year. The
casualties are signiﬁcant in a study of conﬂict resolution. The more destruction,
the more difﬁcult will be peacemaking, reconstruction and the creation of a new
post-war relationship. These distinctions are also relevant from a conﬂict prevention
perspective. It is a common belief that it is in the early phases of a
conﬂict that it can be brought to an end most successfully. Thus, conﬂicts with
lower levels of casualties may reﬂect preventive efforts, not just superiority. It
becomes important to understand which conﬂicts remain on a low level and
which ones do not. The criteria make this possible.
The Uppsala project adds an element which is found neither in COW nor in
AKUF, and it is introduced for theoretical as well as practical reasons. It
requires that the conﬂict should have an issue, an incompatibility. This is
derived from the theoretical considerations that guide this book. In the deﬁnition
of conﬂict given in Section 2.1 this is an important element, and it is, as
a consequence, reﬂected in the data collection. The two other projects are
satisﬁed once they have identiﬁed the actors and the actions. Still, there is an
implicit understanding that only political violence is included. The Uppsala
project handles this openly by requiring that there should be an explicit issue
of contention, deﬁned in political terms. In this way, a clear line is drawn
between political and non-political violence.
The project includes only conﬂicts which concern control over government
or control over territory. These are in turn deﬁned as two exclusive categories.
Control over government means that the issue is who should rule a particular
state, and that demands for change include the change of rulers. The incumbents
are not likely to abide by such a demand easily. Thus, an incompatibility
exists. This means that interventions from abroad to remove a leadership in a
country are recorded as armed conﬂicts (for instance, the United States
intervening in Panama in 1989). So are rebellions against a government by
internal forces (for instance, the uprising against the Mobutu regime in Zaire
in 1997 or against the successor Kabila regime a year later). Control over
24 UNDERSTANDING CONFLICT RESOLUTION
territory means that demands by one state for territory in another state, even
the occupation of another state, are included. So are rebellions inside a state to
achieve autonomy, independence or the joining of a particular territory to the
neighbouring state. This has an international dimension (for instance, Iraq’s
claim on Kuwait, occupying the country in 1990 and being forced out by
1991) as well as an intrastate one (Kosovo Albanians aspiring to leave the
Federal Republic of Yugoslavia, since 1997 expressed also in violent form).
There are theoretical reasons for bringing incompatibility into the conﬂict
deﬁnition. Conﬂict theory suggests that parties act for particular purposes.
Thus, they need initially to be taken at face value. In other analyses such
purposes are regarded as secondary. The Correlates of War project is focused
on armed behaviour, the war. It aims at understanding what triggers this
particular type of behaviour. The project design focuses on structural conditions
as potential explanations, such as balance of power and other elements
in the international system. It does not include the party’s own perception of
why the conﬂict is there. Thus, COW reduces the complexity of the situation
to certain important variables. The same is true for the Hamburg project and
its perspective is equally structural (notably, capitalism or globalization) as are
those of COW (the international system). However, if the focus is shifted to
conﬂict resolution, as is the case for the Uppsala project, the parties’ intentions
become more important. Conﬂict exists, the parties will say, because
there are particular grievances and, thus, the conﬂict cannot end until such
grievances are resolved, ended or at least attended to. With its categories,
the Uppsala project attempts to capture some such basic grievances. This
approach receives interesting support from other studies, pointing to the
importance of territory, for instance, by Holsti (1991) and Vasquez (1993,
1995). The Uppsala project aims at connecting its data to the development of
conﬂict theory, in particular, theories of conﬂict resolution.
There is also a practical consideration, alluded to in the Hamburg project
(Jung et al. 1996: 51), that a line has to be drawn between political violence
and sheer banditry, mutinies and other forms of collective violence. There are
cases where drug dealers clearly are behind the assassination of presidential
candidates. Colombia had such an experience in the 1990s. However, the
purpose is seldom for the assassin and his/her bosses to take control of the
government. The aim is rather to prevent actors from taking power, if they
might affect the government’s policy on drug trade (changes in laws, operations
and effectiveness of the policy). This type of violence is different, as it
reﬂects criminal concerns. Such matters require police strategies, not peace
research. There are delicate borderlines to observe, however. It is known that
regular armed services, grey-zone paramilitary groups as well as many socalled
liberation movements sustain themselves through the drug trade or
other smuggling operations. There are also warlords who draw a thin line
between politics and commerce.
Armed Conﬂicts and Peace Agreements 25
Patterns of armed conﬂict
The number of armed conﬂicts for the period 1989–99 with the Uppsala
deﬁnition is 110. For the eleven-year period, this means that the average
conﬂicts per year is about ten, a much higher ratio than reported in the two
other projects. A comparison made for an overlapping period shows the same,
although not in as striking a manner. For instance, AKUF reports a total of 43
armed conﬂicts in 1994 (as of May that year, Jung et al. 1996: 52–54), while
Uppsala reports 42. Table 2.1 shows the development of armed conﬂict during
the recent period, building on the Uppsala conﬂict data.
Given the discussion on wars, the two bottom rows of Table 2.1 are most
interesting to follow. The trends are not linear. The immediate post-Cold War
period showed an increase in overall conﬂict frequency. The number of about
twenty ongoing wars also parallels what is reported by the Correlates of War
for earlier years. By the middle of the 1990s the total numbers as well as the
numbers of wars declined somewhat. Several wars were brought to a halt or
settled by peace agreements. However, by the late 1990s the number of severe
conﬂicts was again higher.
This pattern is even more pronounced when studying different regions.
Europe, which for a large part of the Cold War saw little manifest armed
conﬂict, was the ﬁrst region to experience a sharp rise in conﬂicts. These were
associated with the break-up of the Soviet and Yugoslav unions. The numbers
went from two armed conﬂicts in 1989 to ten by 1993, by 1997 they were
down to zero, only to see two conﬂicts in 1998 (Northern Ireland and
Kosovo) and three in 1999 (Kosovo, Dagestan and Chechnya). It meant
conﬂicts rebounded by the end of this period in the same volatile regions
(Balkans, Caucasus) as well as in a very protracted conﬂict (Northern Ireland).
For the coming years, the Balkans and the Caucasus stand out as the areas
most in need of conﬂict resolution arrangements and political solutions.
For Africa, there is another pattern. By the middle of the 1990s, this
continent appeared to be a chief beneﬁciary of the end of the Cold War. Wars
on the continent that had been sustained by the Cold War, as well as by South
TABLE 2.1 NUMBER OF ARMED CONFLICTS IN THE WORLD, 1989–99, ANNUALLY
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
Minor 15 16 18 23 15 16 12 17 13 10 10
Intermediate 14 14 13 12 17 19 17 13 14 13 13
Wars 18 19 20 20 14 7 6 6 7 14 14
All 47 49 51 55 46 42 35 36 34 37 37
Source: Uppsala Conﬂict Data Project, Wallensteen and Sollenberg 2000
Note: Among the 110 armed conﬂicts, the highest level reached during the period
was for 50 conﬂicts to be a minor armed conﬂict, for 12 to be on the intermediate
level, whereas 48 reached the level of war. See text for deﬁnitions.
26 UNDERSTANDING CONFLICT RESOLUTION
African polarization, were on the verge of ending. This could be observed for
the Horn of Africa and Southern Africa. West Africa and the Sudan seemed to
be the only conﬂict areas left. From fourteen wars in 1989 and seventeen in
1990 and 1991 the numbers were down to nine in 1995, only to be back to
fourteen by 1998. New wars were experienced in the Horn of Africa (Eritrea
vs. Ethiopia) and, most challenging, in Central Africa, connecting conﬂicts
over a vast, highly populated and resource-rich region. Conﬂicts in this area
became increasingly difﬁcult to disentangle. A regional conﬂict complex was
created, initially centring on the Hutu–Tutsi conﬂict which, by way of refugee
ﬂows, guerrilla movements and interventions, came to engulf a number of
states.1
The high hopes for Africa were dashed, and the optimistic slogan of
‘African Renaissance’ was challenged.
In comparison to the dramatic developments in Europe and Africa, other
regions show a surprisingly permanent pattern of conﬂict. Since the Gulf War,
there has been no other war in the Middle East, albeit armed conﬂict on lower
levels of intensity. The tensions in South Asia between India and Pakistan have
gradually increased, adding a nuclear dimension, and the conﬂict continues to
be active in Sri Lanka as well. The Asian economic crisis was inﬂuential in the
fall of the old regime in Indonesia and may have stimulated efforts to settle the
conﬂict around East Timor. There were tensions between China and Taiwan
as well as on the Korean peninsula, but they are hard to attribute to the
economic crisis. Instead, they repeat a pattern that has been familiar for many
decades. South America has seen a reduction of conﬂicts compared to the
1980s when there were full-scale civil wars going on in Central America and
Peru. These are now either settled or at a low intensity level. Colombia remains
the major exception.
For the world as a whole, the total number of armed conﬂicts going on
is staggering. In spite of great efforts at conﬂict resolution, it appears that
for each conﬂict solved between the parties with international efforts, a new
one emerges, requiring the same mix of improvisation and standard operating
procedures by the international community. This repeated experience of
inadequacy fuels the interest in preventive conﬂict management. Also, it shows
the need to search for the underlying causes in order to ﬁnd remedies that
combine conﬂict prevention with social change and popular participation. In a
longer perspective, none of the armed conﬂicts initiated in the 1990s has been as
devastating as some of the older conﬂicts. The protracted war in Afghanistan,
which began in 1978, has more than 1 million deaths. The 1980s also witnessed
the Iran–Iraq War with 1.2 million deaths. The Vietnam War, or, more appropriately,
the Second Indochina War, ended in 1975, with possibly 2 million
deaths. The Nigerian Civil War in the 1960s was a disaster with 1 million
deaths. The Korean War in the 1950s reportedly led to 2 million deaths. The
civil war in the 1940s in China, ending with the Communist Party taking
control, saw 1 million battle-related deaths (Singer 1991; Small and Singer
Armed Conﬂicts and Peace Agreements 27
1982). The war scenarios since the Cold War are serious enough. Still, it might
be suggested that there is a greater ambition to reduce human suffering. There
are examples of humanitarian support even in the midst of war, in Bosnia, for
instance. Perhaps there is also a willingness from the outside community to act
earlier in a serious conﬂict to prevent it from become more destructive.
2.4 Outcomes of Armed Conﬂict
The concept of conﬂict resolution was given a preliminary deﬁnition in Chapter
1. It will be further reﬁned in Chapter 3. Having delineated armed conﬂict and
war, peacemaking is easier to encircle. It is something done by the warring
parties, expressed in the form of an agreement, implemented by ﬁrst ending
the ﬁghting and then followed through in all other respects. The value of
agreements has been challenged. It is, however, not easy to end a war and start
a process of forging a post-war order without having some shared awareness
of what the new arrangement should look like. An agreement expresses such a
joint understanding. It may not include everything that needs settlement, and
there are likely to develop different interpretations of what has been concluded.
Still, an agreement provides a basis for a new relationship. It is not
easy to make accords after a war, even when the parties have been allies, as
was the case of the USA, Britain and the Soviet Union during the Second
World War. Even before the war ended, the allies had serious disagreements
on the post-war world, although they had made agreements (in Yalta in
January 1945, for instance). If victors cannot agree among themselves, it is
probably even more difﬁcult for enemies to develop a shared document on the
future of their interaction. A peace agreement, in other words, is a particular
result in a process that began before the agreement was signed and continues
after the ink has dried.
The Uppsala project is collecting data on peace agreements. It makes possible
closer scrutiny of frequencies of different types of war ending. For instance,
for the period 1989–99, with a total of 110 armed conﬂicts, it was recorded
that 75 had actually been terminated by the end of 1999, that is more than
two-thirds of the total (Wallensteen and Sollenberg 2000). This supports the
statement in section 2.1 that conﬂicts and wars actually do end.
The endings vary, however. Armed conﬂicts may, according to the Uppsala
project, end in (1) victory, meaning the capitulation of one side to the other in
the form of an agreement, a withdrawal from the battle or in other ways. The
conﬂict may also (2) continue at such a low level of action that it does not
reach the threshold of 25 battle-related deaths. It might return after some time
which means that what was recorded was only a lull when the parties were
recuperating. In some cases of little recorded violence, there might be a
clandestine capitulation agreement, a hidden peace treaty or a cease-ﬁre
28 UNDERSTANDING CONFLICT RESOLUTION
arrangement. This is deﬁned as ‘other outcomes’. However, the most interesting,
in view of our discussion, is (3) the peace agreement, openly acknowledged,
negotiated between the parties and explained by them to their own
followers and public. Finally, at any moment in time, some conﬂicts might still
be ongoing, constituting an additional category (4) of non-terminated conﬂicts.
Placing the 110 conﬂicts in these categories, Table 2.2 emerges, also
displaying a regional breakdown of the outcomes. Note that the three columns
to the left add to make the fourth column. Adding this column to the ﬁfth one
gives us the total number of conﬂicts in the region for these eleven years.
In Table 2.2 it can be seen that peace agreements are almost as frequent an
outcome as are victories, and other outcomes. Victory, in fact, does not occur
in more than around one-ﬁfth of all conﬂicts going on in this period.
Table 2.2 shows that the outcomes vary and that there are no simple
explanations for one type of ending. All regions have experienced a mixture of
victory, defeat, peace agreements and other outcomes. There is a possible
tendency that conﬂicts in Asia and Africa are terminated less quickly than
those in Europe, for instance, but the meaning of this observation is not clear.
It is safer to say that the variety of outcomes often is not what the initiators
have normally expected. The preference is for victory within a reasonably
short period. However, the largest single category is actually the one of
continued conﬂicts. Many conﬂicts are deeply entrenched, have witnessed
broken negotiations, failed cease-ﬁre arrangements, and abandoned peace
agreements. They are probably increasingly difﬁcult to settle. For many
initiators, however, what was originally planned may no longer be possible to
accomplish. One of the most protracted conﬂicts is the one in Afghanistan.
The war began as an attempt by a Communist Party to reform the feudal
society, change the land distribution and give women a stronger standing.
After twenty years of war with many special features – Soviet invasion, US
support to opposition movements and involvement from a host of neighbouring
and Middle Eastern countries – the Communist Party was eliminated and
many of its leaders brutally murdered. The conﬂict was then pursued along
TABLE 2.2 OUTCOMES OF ARMED CONFLICTS, 1989–99, BY THE END OF 1999
(1)
Victory
(2)
Other
Outcomes
(3)
Peace
Agreements
(1)؉(2)؉(3)
Total
Terminated
(4)
Not
Terminated All
Europe 8 7 4 19 2 21
Middle East 3 5 1 9 2 11
Asia 2 10 5 17 13 30
Africa 3 9 7 19 16 35
The Americas 6 1 4 11 2 13
All regions 22 32 21 75 35 110
Source: Uppsala Conﬂict Data Project
Armed Conﬂicts and Peace Agreements 29
traditional divisions, and the dominating group until November 2001 (the
Taliban) was unusually Islamic and anti-women. The conﬂict dynamics have
become entirely different from what the originators had anticipated.
Victory is difﬁcult to achieve. It does occur, however. The most obvious
example is the USA intervening in Panama, capturing the ‘strongman’ of the
leadership, General Noriega, bringing him to trial in Florida, convicting him
for drug trade offences and putting him in a US prison for 30 years. The Gulf
War is also a victory: Iraqi forces had to withdraw, and Kuwait was restored
as a sovereign country. In 1997 a rebellion against the incumbent regime in
Zaire ended with victory. The war lasted eight months. The new regime faced
another rebellion less than nine months later. A peace agreement concluded in
July 1999 and signed in September the same year, was meant to end the war,
without accomplishing this by the end of 2000.
The dynamics of victory and defeat are known from history. The large
number of peace agreements is a more novel aspect, and part of the experience
since the Cold War. Table 2.2 includes 21 such agreements. There are additional
agreements. For instance, there are accords concerning wars that went
on before 1989 (for instance, Chad vs. Libya 1990, Israel vs. Jordan 1994,
South Africa with Angola and Namibia 1988, implemented in the years
thereafter). Also there are treaties in conﬂicts that were limited (Central
African Republic in 1997). Furthermore, some agreements outlined processes
towards a solution (Israel vs. Palestine since 1993) where other parties nevertheless
pursued a violent ending. There are even cases of a complete settlement
concluded between the main parties, where other actors took up or continued
armed struggle nevertheless (Mindanao in 1996). Of course, there are also
agreements that have been functioning for a period of time, but then have
been undermined by the parties. This is true for the settlements for Angola in
1991 and 1994, Chechnya in 1996 and Sierra Leone 1996 and again in 1999.
The record of successful peacemaking is as varied as can be expected from the
difﬁculties of ending long-lasting wars. Still, the ambition to do so with the
help of negotiation and agreement makes the period since the end of the Cold
War an interesting object of study.
The developments of the post-Cold War period can legitimately be compared
to the very few peace agreements ending wars that were concluded
during the entirety of the Cold War. There were cease-ﬁre agreements, no
doubt, but very few peace agreements. As we noted already in Chapter 1,
some of the cease-ﬁre lines drawn during the Cold War still constitute the
main territorial divisions in many conﬂicts. To the cases mentioned previously
we can also add the territorial division between India and Pakistan in Kashmir
in 1949 that ended their ﬁrst war. New wars in the area in 1965 and 1971 led
to a return to the previous cease-ﬁre lines. The conﬂict remains unresolved.
Among the few real peace agreements concluded during the Cold War, the
Geneva peace agreement for Indochina in 1954 was effectively undermined
30 UNDERSTANDING CONFLICT RESOLUTION
within two years. The war between Malaysia and Indonesia, which began in
1963, was concluded through a peace pact in 1966, and the conﬂict has not
resumed. A short war in 1963 between Algeria and Morocco found a mediated
agreement the same year. An agreement in 1972 to end the war in the
Sudan was shattered in 1983 and war returned. The Camp David agreement
between Israel and Egypt in 1979 has stood the test of time. For the 45 years
of Cold War, the peace agreements are few, probably not more than ten if we
apply the deﬁnition introduced here (Licklider 1995; Mason and Fett 1996;
Stedman 1991). This is not a particularly striking record at peacemaking. It
contrasts the many arms control agreements made, where one source lists 27
international accords from 1963 to 1991 (Goldstein 1992). This also makes
the large number of peace agreements during the turbulent period of the
1990s valuable and worth a close analysis.
Table 2.2 demonstrates that peace agreements have been concluded in all
regions of the world. This means that peacemaking has taken on a global
meaning. Without the peacemaking efforts the number of wars would probably
increase signiﬁcantly. The agreements may have been concluded between
parties too exhausted to ﬁnd resources to win the wars, but sometimes also
not capable of concluding agreements on their own (The Dayton Agreement
on Bosnia-Herzegovina may ﬁt in this category, as the warring parties only
accepted the deal under strong international pressure). Thus, we will proceed
into the theme of conﬂict resolution by departing from the peace agreements
concluded or implemented since 1989. With this in mind, it is ﬁrst necessary
to turn to a more theoretical discussion on conﬂict theory and its implications
for conﬂict resolution. This is done in Chapter 3.
Armed Conﬂicts and Peace Agreements 31