|
Patterns of
conflict
occurrence
Causes and processes of
conflict
Costs of conflict and
conflict early
warning
|
|
Is the world
more or less violent today than in the past? Are wars more or less
destructive than they used to be? Are modern conflicts different from
earlier ones? What are the causes of conflict initiation, continuation
and termination? The systematic study of violent conflicts seeks
answers to these and related questions. This web page presents the
world's primary English-language data-collection projects. It reflects
the efforts of a field of study that has become increasingly diverse
and complex, since Quincy Wright first published his pioneering work A
Study of War in 1942. The
16 data sets described below are grouped according to whether they
focus primarily on the
patterns of conflict occurrence, the
causes and processes of conflict, or costs
of conflict and conflict early warning.
The
categories overlap and are intended only as a rough guide. This list is
not comprehensive, but an attempt has been made to make it complete
within certain parameters. Every data project is directly concerned
with conflict, provides worldwide coverage, is publicly available in
English and is widely judged to be reputable. A large number of data
sets created for specific publications and made available by authors
are not included here because they constitute data use rather than data
collection. For a discussion of the the theoretical, methodological and
policy-related questions that face researchers in the systemic study of
conflict, please refer to Appendix 1C:
Measuring Violence: An Introduction to Conflict Data Sets in the
SIPRI Yearbook 2002.
Patterns of Conflict Occurrence
Location:
University of Michigan, Ann Arbor, Michigan, USA, and Pennsylvania
State University, State College, Pennsylvania, USA.
Principal investigators: J. David Singer, University of
Michigan; and Stuart Bremer, Pennsylvania State University.
Purpose:
To promote and support the scientific study of the causes of war and
the conditions of peace by collecting and processing large quantities
of historical information in an attempt to identify and explain
empirical regularities that lead to war. Interstate conflict is the
special focus of the project, with emphasis on those conflicts that
involve the threat, use or display of force. Intra-state and
extra-systemic conflicts are also studied.
Current coverage: 1816–1997.
Dependent variables:
(a) Interstate war is sustained combat between the regular military
forces of two or more state members of the international system in
which there is a total of at least 1000 battle-related fatalities. (b)
Intra-state war is sustained armed combat between two armed forces
within the boundaries of a state, in which there are at least 1000
battle-related fatalities per year. (c) Extra-systemic war is sustained
armed combat between a state member of the international system and a
non-system member political entity outside its territorial boundaries,
in which there are at least 1000 battle-related fatalities per year. Earlier
data sets, code books, publications and contact information are
also available.
Back to the top
Location:
Department of Peace and Conflict Research, Uppsala University, Uppsala,
Sweden.
Principal investigators: Peter Wallensteen and Margareta
Sollenberg.
Purpose:
To collect information on selected variables relating to armed
conflict, primarily to be used in research on various aspects of the
origins, dynamics and resolution of conflict. Data have been collected
on a global and yearly basis since 1989.
Current coverage: 1989–2001. The project has recently
collaborated with others to extend the coverage from 1946 to 1988.
Dependent variable:
Armed conflict is a contested incompatibility that concerns government
or territory or both, over which the use of armed force between the
military forces of two parties results in battle-related deaths. At
least one of the parties is the government of a state. (a) Minor armed
conflict results in at least 25 deaths per year and fewer than 1000
deaths over the course of the conflict. (b) Intermediate armed conflict
results in more than 1000 deaths during the course of the conflict, but
fewer than 1000 in any given year. (c) War results in more than 1000
deaths in any given year. The data set covering the extended period of
1946–2001 will be available in late 2002 on the Internet site of the International
Peace Research Institute, Oslo (PRIO).
Note: data from the Conflict Data Project have been published in the SIPRI Yearbook
since 1987.
Back to the top
(Konflikt-Simulations-Modell,
KOSIMO)
Location:
Heidelberg Institute of International Conflict, University of
Heidelberg, Heidelberg, Germany.
Principal investigator: Frank R. Pfetsch.
Purpose:
To provide a searchable database of political conflicts including
crises, wars, insurrections, negotiation, mediation and peace
settlements.
Current coverage: 1945–99.
Dependent variable:
Political conflict is defined as the clashing of overlapping interests
around national values and issues between at least two parties, at
least one of which is the organized state. The conflict has to be of
‘some duration’ and ‘magnitude’. The intensity of conflicts ranges from
‘latent conflict’ to ‘non-violent crisis’ to ‘violent crisis’ to ‘war’.
Possible instruments used in the course of a conflict are negotiations,
authoritative decisions, threat, pressure, passive or active
withdrawals, or the use of physical violence.
Back to the top
Location:
Center for Systemic Peace, University of Maryland, College Park,
Maryland, USA.
Principal investigator: Monty G. Marshall.
Purpose:
To list all episodes of major political violence of any type.
Categories include all forms of inter-state, intra-state, and
inter-communal warfare. This data set is one of six data sets that
comprise the Armed Conflict and Intervention Project at the Center for
Systemic Peace and the Center for International Development and
Conflict Management (CIDCM), University of Maryland. The larger project
attempts to capture the deeper qualities and complexities of violent
social conflict. It collects global information on the security
context, membership of international organizations, displaced
populations, direct military interventions, political interactions and
bilateral trade flows.
Current coverage: 1946–2000.
Dependent variable:
Major episodes of political violence involve the systematic use of
lethal violence and terror by organized groups and/or states that
substantially affect the society or societies that directly experience
the armed conflict (resulting in at least 500 directly related
fatalities, substantial destruction of infrastructure and population
displacements). Episodes may involve states, a state and non-state
group, or non-state groups only, including inter-state and independence
war, ethnic and revo-lutionary (civil) war, inter-communal warfare,
genocide and communal massacres. Each episode is rated on a ten-point
scale according to its total impact on the society or societies that
are directly affected by the violence.
Causes and Processes of Conflict
Location:
University of Maryland, College Park, Maryland, USA and McGill
University, Montreal, Quebec, Canada.
Principal investigators: Michael Brecher, McGill
University and Jonathan Wilkenfeld, University of Maryland.
Purpose:
To investigate 20th century interstate crises and the behaviour of
states under externally generated stress. The data describe the
sources, processes and outcomes of all military–security crises
involving states.
Current coverage: 22 December 1917 to 31 December 1994.
Dependent variable:
Part 1: All international crises occurring during the coverage period,
characterized by: (a) a distortion in the type and an increase in the
intensity of disruptive interactions between two or more adversaries,
with an accompanying high probability of military hostilities or,
during a war, an adverse change in the military balance; and (b) a
challenge to the existing structure of an international system—global,
dominant or subsystem—posed by the higher-than-normal conflictual
interactions. Part 2: All foreign policy crises experienced by states
due to their involvement in the international crises defined above. A
foreign policy crisis is defined as a situation in which three
conditions, deriving from a change in a state’s external or internal
environment, are perceived by the highest-level decision-makers of the
state: (a) a threat to basic values, (b) an awareness of finite time
for response to the external threat to basic values, and (c) a high
probability of involvement in military hostilities.
Location:
Pennsylvania State University, State College, Pennsylvania, USA.
Principal investigators:
Stuart Bremer, Jim Ray, Dan Geller, Paul Diehl, Doug Gibler, Paul
Hensel, Chuck Gochman, Glenn Palmer, Brian Pollins, Ric Stoll, Pat
Regan and Zeev Maoz.
Purpose: To identify for all
militarized interstate disputes the participants, start and end dates,
fatality totals, hostility levels, revision sought, outcome and method
of settlement.
Current coverage: 1816–1992. Currently being updated to
2001.
Dependent variable:
A militarized interstate dispute involves the threat, display or use of
force short of war by one member state, explicitly directed towards the
government, official representatives, official forces, property or
territory of another state. The outcome variable is recorded on a
five-point ordinal scale ranging from non-reciprocated action, threat,
display, use of force, to interstate war.
The data span from 1816–1992, an update
to 2001 will be available from late 2002.
Back to the top
Location:
Middlebury College, Middlebury, Vermont, USA.
Principal investigator: Russell Leng.
Purpose:
To analyze the behaviour of states engaged in potential pre-war
disputes. Data on 47 cases focuses on crisis dynamics by generating
descriptive data on the actions and interactions of states. The cases
are selected as a representative sample of crises since 1816 and they
permit the testing of a number of theories of interstate crisis
behaviour.
Current coverage: 1838–1980.
Dependent variable:
Interstate crises in which the principal protagonist on each side is a
member of the interstate system. A crisis is an event on the continuum
of belligerence that extends from a simple dispute, to a militarized
dispute, to a crisis, to war. A crisis is a militarized dispute which
requires protracted bargaining, defined as when the two major
participants exchange at least 50 acts.
Back to the top
Location:
Florida State University, Tallahassee, Florida, USA.
Principal investigators: Paul Hensel and Sara McLaughlin
Mitchell.
Purpose:
To collect systematic data on contentious issues between states, with a
focus on identifying the issues regardless of any particular action
that may or may not have been taken to resolve them. Presently covering
territorial, riverine and maritime claims.
Current coverage: 1816–2000.
Dependent variable:
Contentious issues which involve explicit statements of disagreement by
official governmental representatives of at least two states. ICOW
identifies each issue with reference to the involved states, the object
of the claim (such as the specific river or territory) and the time
frame over which it endures. ICOW then collects data on the salience of
each issue and on attempts to settle it through peaceful bilateral
negotiations, binding or non-binding third-party activity or
militarized conflict.
Back to the top
Location:
Department of Political Science, University of Illinois, Urbana,
Illinois, USA.
Principal investigators: Paul Diehl and Gary Goertz.
Purpose:
This data set provides a comprehensive overview of 1166 rivalries, 63
of which are enduring. The data set is designed to provide the basis
for the analysis of the initiation, dynamics and termination of
international rivalries.
Current coverage: 1816–1992.
Dependent variable:
Rivalry is defined by the frequency of militarized interstate disputes
between the same pair of states. The existence of a mili-tarized
rivalry is indicated by the occurrence of militarized disputes as
defined by the COW–MID data set. Disputes which occur within 10–15
years of each other are con-sidered to be part of the same rivalry. A
dispute is considered part of the same rivalry if it involves the same
two states and occurs within 11 years of the first dispute of the
sequence, 12 years after the second dispute, and up to 15 years after
the fifth dispute.
Back to the top
Location:
Center for International Development and Conflict Management (CIDCM),
University of Maryland, College Park, Maryland, USA.
Principal investigators: Monty G. Marshall, Ted Robert
Gurr, Jack A. Goldstone and Barbara Harff.
Purpose:
To provide the dependent variable (state failure) in the US Government–
sponsored State Failure Task Force quantitative analyses of structural
indicators of failure, the purpose of which is to create ‘early
warning’ (two-year) models of state failure situations. Independent
variables used in published Task Force analyses are being prepared for
public release and should be available soon to complement the dependent
variable data on state failures.
Current coverage: 1955–2000.
Dependent variable:
The data set includes all cases of internal wars (ethnic war,
revolutionary war, or genocide–politicide) and failures of governance
(substantial reversion to more autocratic rule or collapse of central
authority) that began between 1955 and 2000 in independent countries
with populations greater than 500 000. A war is defined as an armed
conflict involving state authorities and a challenger group that
results in at least 1000 directly related deaths over the course of the
episode and at least one year during which there was more than 100
directly related deaths. The case begins with the first year during
which the 100 death threshold is reached and ends when deaths fall
below that threshold for at least five years. An episode of genocide or
politicide (politically motivated mass murder) is defined by the merits
of the case (for instance, an established intent to eliminate
non-combatant group members). A failure of governance is defined
generally as a six-point decrease in the state’s Polity IV regime score
(that is, towards greater institutional autocracy) or a Polity IV
‘interregnum’ (a collapse of central regime authority through failure,
revolution, or involuntary state disintegration).
Back to the top
Location:
Center for International Development and Conflict Management (CIDCM),
University of Maryland, College Park, Maryland, USA.
Principal investigators: Ted Robert Gurr, Monty G.
Marshall and Christian Davenport .
Purpose:
To monitor and analyse the status and conflicts of politically active
ethnopolitical groups in countries with a population of at least 500
000. Coverage of 275 contemporary and 65 historical groups.
Current coverage: 1945–2000.
Dependent variable:
Ethno-political groups, defined as communal groups that: (a) are
disadvantaged by comparison with other groups in their societies,
usually because of discriminatory practices, or (b) have organized
politically to promote or defend their collective interests. Only
ethno-political groups with populations greater than 100 000 or 1 per
cent of the population are included.
Back to the top
Location:
University of Indianapolis, Indianapolis, Indiana, USA.
Principal investigator: Bill Ayres.
Purpose:
To measure and study conflict outcomes in violent, intra-state
nationalist conflicts—those involving ethnic and other forms of
secessionism—and their antecedents and correlates. It seeks to mark
starting and ending points of these conflicts, and measure the
characteristics and behaviours of the actors in them, to answer
questions about how and why conflicts end the way they do.
Current coverage: 1945–96.
Dependent variables:
Each conflict episode is coded for the highest level of violence or
rebellion reached during that episode, using a seven-point ordinal
scale adapted from the Minorities at Risk data set project. Each
conflict episode is coded for estimated number of deaths caused by that
conflict (rounded to the nearest 100 for estimates less than 10 000,
rounded to the nearest 1000 for estimates greater than 10 000).
Conflict ends when both sides are no longer either fighting or talking
with each other about what the solution to the conflict should be.
Different episodes of the same conflict must be separated by a 12-month
lull in both fighting and negotiating. The data also specify four types
of ending and four types of agreement.
Back to the top
Location:
University of Binghamton, Binghamton, New York, USA.
Principal investigator: Patrick M. Regan.
Purpose:
To identify all military, economic and diplomatic interventions into
civil conflicts, primarily to help determine the relationship between
military, economic and diplomatic interventions and the duration of
conflict. Current coverage of military and economic interventions.
Current coverage: 1945–99.
Dependent variable:
Third-party interventions in intra-state conflicts are
convention-breaking military and/or economic activities in the internal
affairs of a foreign country targeted at the authority structures of
the government with the aim of affecting the balance of power between
the government and opposition forces. Intrastate conflicts are
organized military hostilities between two groups in conflict in which
there were at least 200 fatalities over the course of the conflict.
Back to the top
Costs of conflict
and conflict early warning
Location:
Program on Nonviolent Sanctions and Cultural Survival (PONSACS),
Harvard University, Cambridge, Massachusetts, USA.
Principal investigators: Doug Bond, Joe Bond, J. Craig
Jenkins, Churl Oh and Charles Louis Taylor.
Purpose:
PANDA is an automated early warning system that is combined with
on-the-ground research of conflict regions provided by anthropological
insights. These two strands of research at PONSACS work to identify
conflict regions before they erupt into violence and to actively
promote nonviolent alternatives to armed conflict. The project’s
premise is that by monitoring and examining interaction events with a
‘data lens’ that is sensitive to nonviolent direct action, it can track
and compare the evolution of conflict manifest in both violent and
nonviolent behaviours. The project also seeks to help make the costs of
conflict transparent by providing a longitudinal series of social,
political and economic events gleaned from the leads of news reports,
and to facilitate independent testing and peer review. The PANDA
software protocol for parsing news stories has been superseded by the
Integrated Data for Events Analysis (IDEA) protocol.
Current coverage: 1991–2000.
Dependent variable:
The basic parameters of the data include the source actor and target
actor of social, political and economic events, the events themselves,
as well as their date, location and a selection of attributes of the
same. In more common terms, each data record represents the ‘who does
what to/with whom, when, where, why and how’ of an event reported in
the news. Any of the events data variables may be treated variously as
an independent or dependent variable, depending upon the specific
research questions being asked.
Updated
data using the IDEA protocol are also available. The site provides
database access, a user’s manual, and generation of data sets and
summary tables. Access to the database requires a password and its use
is restricted to non-commercial, academic use.
Back to the top
Location:
Center for International Development and Conflict Management (CIDCM),
University of Maryland, College Park, Maryland, USA.
Principal investigator: John Davies.
Purpose:
To allow computer-assisted identification, narrative description and
analytical coding of daily international and intra-national events,
describing the day-to-day actions of all states and the major non-state
communities and international organizations. GEDS includes and expands
on Edward Azar’s Conflict and Peace Data Bank (COPDAB, covering
1948–78), updating it for selected countries. Near-real-time tracking,
as needed for early warning, can be generated on request using COPDAB
scales or more specific ‘accelerator’ models for anticipating
ethno-political conflicts, genocides or politicides.
Current coverage: 1948 to the mid-1990s.
Dependent variable:
Events are operationally defined as reports from reputable sources
which specify who did or said what to whom, when and where. Conflict
and cooperation are operationally defined and coded using Azar’s
15-point COPDAB scales either as categorical variables, as ordinal
variables, or as ratio variables.
Back to the top
Location:
University of Kansas, Kansas City, Kansas, USA.
Principal investigator: Philip A. Schrodt.
Purpose:
To generate political event data through automated coding of
English-language news reports. These data are used in statistical
early-warning models to predict political change. Building on the World
Event/Interaction Survey (WEIS) Project, the KEDS project has three
major research concentrations: software development for the
machine-coding of political event data, production of events data sets
and development of early-warning methods.
Current coverage: 1979–99.
Dependent variable:
Event data are nominal or ordinal codes recording the reported
interactions between international actors at specific points in time.
|
|
Introduction
Patterns of Major Armed Conflicts
Conflict Data Sets
Conflict Prevention
|
New SIPRI paper on Ukraine's small arms and light weapons exports
