Methodology of the Heidelberg Conflict Research
Conflict research in Heidelberg spans the work of two organizations located at the Institute for Political Science of the University of Heidelberg: the Heidelberg Institute for International Conflict Research (HIIK), founded in 1991, and the research group Conflict Information System (CONIS Group), established in 2005.
Quantitative conflict research as basic research follows the principal goal of empirically documenting conflict events as precisely as possible, aiming at providing rich data sets for further research on the causes of conflicts. However, comparative meta-analyses have shown that the existing conflict data sets only marginally share the same coverage. Alongside factors such as the scope and reliability of the data and its acquisition, a further substantial factor can explain the observed discrepancies between the various databases, namely the difference between methodologies employed to extract and assess data, specifically concerning the concept and operationalization of the conflict definition.
In the field of quantitative conflict research, one may essentially find three different approaches to conceptualizing and operationalizing a definition of a conflict or war:
(1) Official war declarations. This approach was used in early attempts to quantitatively document interstate conflicts. Meanwhile, this understanding is seen as outdated, for contemporary wars are normally not being officially declared anymore, partially due to norms of international law.
(2) Number of casualties. This method of quantitative operationalization is being used by the two most prominent conflict databases, the Correlates of War Project (COW), and the Uppsala Conflict Data Program (UCDP). Both projects regard the threshold of 1,000 conflictrelated deaths as well as the participation of at least one state as the most important criteria for the existence of a war. In order to be able to document smaller violent conflicts, UCDP created the Minor Armed Conflicts category. The threshold for this category is 25 casualties per year.
(3) Mode and quality of the course of the conflict. The Hungarian peace and conflict researcher Istvan Kende presented a further method, in which he took the quality of the development of a conflict as the basis for constructing a concept of wars. The latter approach was adopted by two German conflict-database projects, the Working Group for Research on the Causes of War (AKUF) in Hamburg and the Conflict Simulation Model (KOSIMO) in Heidelberg, both founded independently from each other in the late 1980s.
The latter approach was adopted by two German conflict-database projects, the Working Group for Research on the Causes of War (AKUF) in Hamburg and the Conflict Simulation Model (KOSIMO) in Heidelberg, both founded independently from each other in the late 1980s. Both the HIIKs Conflict Barometer and the CONIS database are methodologically consistent with the qualitative tradition of conflict definition. Extending beyond the approach of AKUF and the KOSIMO database, from which the HIIK has emerged and which is considered the predecessor of CONIS, Heidelberg conflict research, with CONIS in particular, additionally records conflict-related measures and events, prompting the assessment of conflicts and the designation of according intensities. It is only by following this procedure that the resulting evaluation of conflicts becomes inter-subjectively comprehensible and verifiable.
To that end, up until 2010, conflict-measures served as significant indicators, occasionally adding revised casualty and refugee numbers as further indicators. Following a methodological revision in 2011, changes in the Heidelberg methodology included allocating intensities not only based on state units and calender years, but also on the level of subnational units and calender months. Furthermore, the establishment of intensities now follows an analysis of clearly conceived proxy indicators, used for the assessment of means and consequences of a conflict measure. This analysis continues to be based on the actions of conflict actors, as well as the communication between them. Through the conceptual refinement and standardization of data collection and documentation, the HIIK data achieves a higher degree of precision, reliability, and comprehensibility regarding information on political conflicts.
In the following sections the basic concept of a political conflict shall be introduced, thereafter clarifying in detail the definitions and operationalization of the various conflict intensities.
1. The Basic Concept of a Political Conflict
A political conflict is a positional difference between at least two assertive and directly involved actors regarding values relevant to a society (the conflict items) which is carried out using observable and interrelated conflict measures that lie outside established regulatory procedures and threaten core state functions, the international order, or hold the prospect of doing so.
According to the Heidelberg methodology, the essence of a political conflict lies in a contradiction, adequately represented by the concept of a positional difference: a positional difference is a perceived incompatibility of ideas and beliefs. It presupposes the presence of the following elements:
(1) There must be at least two entities possessing intellectual capacity and vision, and who are capable of communicating. Such an entity is called an actor.
(2) In order for the actors to sense incompatibility between their ideas and beliefs, there must be reciprocal actions and acts of communication between said actors. These actions and acts of communication are called measures.
(3) A communicating act always refers to a specific issue, an action always refers to a certain object. The subject behind a measure is called item.
For the purpose of defining the term political conflict more precisely, the three elements aforementioned shall be further defined. These elements are necessary requirements for the existence of a political conflict.
1.1 Defining Conflict Actors
Conflict actors are individual or collective actors who (a) are conceived as unitary, distinguishing themselves from one another through their internal cohesion and internally shared goals, and (b) are perceived as assertive.
A conflict actor who claims the conflict item for itself is called a directly involved or direct actor. A conflict actor who does not claim the conflict item for itself but communicates and acts with regard to the item is called an indirectly involved or indirect actor.
An indirect actor who supports a direct actor in claiming or aspiring the conflict item is called a supporter.
An indirect actor who wishes to end the conflict without supporting any of the direct actors in their actions regarding the item, is called an intervener.
A conflict actor is considered assertive when it is perceived as such by a direct conflict actor, hence causing a behavioral change concerning the conflict item in the latter. The Heidelberg conflict-research method includes both state and non-state actors (private, individual, and collective actors, as well as supranational and international organizations).
1.2 Defining Conflict Measures
Conflict measures are described as all actions and communications by a direct or indirect actor within the context of a specific political conflict. There are two types of conflict measures: constitutive conflict measures, the presence of which establish or sustain the existence of a political conflict on a certain intensity level, and corollary conflict measures, appearing alongside constitutive conflict measures.
A constitutive measure is given when the action or communication in question (a) lies outside established regulatory procedures, and (b) threatens, possibly in combination with other conflict measures, core state functions or the international order, or at least holds the prospect of doing so. Constitutive conflict measures can be violent or non-violent.
A measure lies outside established regulatory procedures when there is either no established procedure for the settlement or elimination of an existing positional difference between conflict actors, or when the measure itself constitutes such a regulatory procedure, but over the existence and manifestation of which the conflict actors disagree. A regulatory procedure can only be seen as established as long as it works to prevent violence, that is, as long as its implementation does not contain actual physical violence or the threat of such.
Core state functions constitute the maintenance of (a) the security of a population, (b) the integrity of a territory and (c) a socioeconomic or cultural order. A core state function or the international order is at risk if, in one of the conflict actors perception, the continuation of conflict measures makes the performance of said functions or the maintenance of the international order impossible or unlikely.
Corollary conflict measures are such actions or communications, which are carried out by a direct actor as part of an established political conflict, but either (a) lie within established regulatory procedures or (b) do not threaten core state functions and/or the international order, or both. Said actions or communications are primarily non-violent conflict measures carried out in conflicts that can be of a violent level.
The HIIK methodology differentiates between conflict measures and conflict events. Conflict events are actions, communications, or situations occurring in the natural surrounding world which are not carried out by one of the conflict actors but influence the development of a conflict.
1.3 Defining Conflict Items
Conflict items are understood as material or non-material goods which are claimed by the direct conflict actors through constitutive conflict measures. Since constitutive conflict measures threaten a core state function, the international order or hold the prospect of doing so, the conflict items attain social relevance. Hence the item bears a relation to the social community within a state or the coexistence of states within the international system. In particular, the Heidelberg methodology allows for the following conflict items depending on which type of good is claimed by the conflict actor.
- Ideology/System: change of the ideological, religious, socioeconomic or legal orientation of the political system or a change of the type of regime.
- National Power: control of a state.
- Autonomy: achievement or extension of political autonomy of a group in a state or of a dependent region without claiming independence.
- Secession: separation of part of a state’s territory with the aim of creating a new state or the incorporation to an existing state
- Decolonization: independence of a dependent territory.
- Subnational Predominance: de facto control of a government, a non-governmental organization or a population over a territory or a population.
- Resources: ownership of natural resources or raw materials, or the profit generated thereof Territory: change of the line of an interstate border
- International Power: changes in the relations of power in the international system or in one of its regional systems.
- Other: residual category.
Conflict actors can claim multiple items at the same time. In addition, different conflict actors can claim different goods, respectively.
2. Definition and Operationalization of the Conflict Intensities
2.1 Definition of Intensity Levels
The conflict intensity is a feature of all conflict measures in a geographical and temporal place. The primary temporal unit of analysis is the calendar month; the primary geographical unit of analysis is the region, that is the top-level, subnational, political division of the state.
The HIIK methodology distinguishes between five levels of intensity: dispute, non-violent crisis, violent crisis, limited war, and war. The category of violent conflicts include the violent crisis, the limited war and the war level, other than the nonviolent conflicts (dispute and non-violent crisis), which are referred to as low intensity conflicts. Accordingly, the violent crises are medium intensity conflicts, while wars and limited wars are high intensity conflicts. (see Figure 1)
Figure 1: Levels of Conflict Intensities
A political conflict is classified as a dispute if it meets all elements of the basic concept.
A political conflict is classified as a non-violent crisis if physical violence is being implicitly or explicitly threatened to persons or property by at least one of the actors, or if one actor uses physical violence against property, without regarding the injury of people as acceptable. A threat of force is understood as a verbally or non-verbally communicated prospect of violent measures that is based on the conflict item. An acceptance of the injury of persons is present when the use of force against property includes the possibility of the injury of people but the violator is indifferent to this possibility.
A political conflict is classified as a violent crisis when at least one actor uses force sporadically against persons – or things in case that physical violence against people is considered acceptable. The applied means and consequences altogether are limited.
A political conflict is classified as a limited war when at least one actor uses force against persons and maybe things in a distinctive way. The applied means and consequences altogether are serious.
A political conflict is classified as a war when at least one actor uses force massively against persons and maybe things. The applied means and consequences altogether need to be framed as extensive.
2.2 Operationalizing Violent Conflicts
To subdivide conflicts into levels of intensity, several criteria have to be taken into account, the first of which is the differentiation of non-violent and violent conflicts. Violence, in this context, is defined as the deliberate exertion of physical injury. Intensities can be generally divided into (a) an absolutely non-violent approach to resolve a conflict, (b) threats of the use of force against persons, and (c) the direct use of force against persons.
- An absence of violence is classified with the intensity level 1.
- Intensity level 2 comprises both cases in which an actor threatens violence and cases that involve damage to property as long as the actor is not indifferent to the possibility that people are being physically harmed.
- If an actor uses force against persons or property while the actor is indifferent to the possibility that persons might be harmed, the conflict is classified with the intensity levels 3, 4, or 5.
Other criteria to further subdivide violent conflicts into levels 3, 4 and 5 are means and consequences applied in a violent conflict measure. Indicators of the category means are the use of weapons and personnel, while indicators of the category consequences comprise casualties, damage, and the number of refugees.
2.2.1 Means of Applied Violence – Weapons and Personnel
To measure the applied means of violence used by all involved conflict actors it is necessary to account for the amount of used material (weapons) as well as the number of involved individuals (personnel).
188.8.131.52. Indicator Weaponry
The weaponry used in a conflict can be divided into two groups, light weapons and heavy weapons:
|Light Weapons||Heavy Weapons|
|Melee weapons||blade weapons (e.g. sword, knife, axe, machete, etc.)
pole weapons (e.g. lance, spear, etc.)
unconventional (e.g. club, bludgeon, baseball bat, etc.)
|Bows and throwing weapons||bow and arrow, crossbow, slingshot, spear, throwing knife, throwing axe, bola, boomerang, throwing star
unconventional (e.g. stones etc.)
|Small arms||pistol, revolver, rifle, carbine, shot gun, submachine gun, assault rifle, light and heavy machine gun, etc.||shoulder-fired anti-tank and anti-aircraft systems, grenade launchers with armor-piercing ammunition|
|Explosive weapons||hand grenade, anti-personnel mine, directional mine, explosive charge, Molotov cocktail, etc.||large conventional bombs, anti-tank mines, conventional missiles, torpedos, depth bombs, naval mines, cruise missiles, etc.|
|Incendiary weapons||flame thrower, Molotov cocktails, etc.||incendiary bombs (napalm, phosphorous), etc.|
|Heavy weaponry||battle tank, infantry fighting vehicle, attack helicopter, fighter jets, battle ships, submarines, etc.
heavy mortars, Maschinenkanonen, anti-tank and anti-aircraft cannons, heavy rocket launchers, naval artillery, artillery, etc.
|ABC weapons||nuclear, biological and chemical weapons (weapons of mass destruction) and their launch facilities|
Table 1: Examples for Light and Heavy Weapons
Heavy weapons again are differentiated by the way they are used. They can be used for both a powerful and a light exertion of violence. The use of a weapon can be either limited or extensive, the former characterizing a light use of a heavy weapon and the latter a powerful use of a heavy weapon.
The evaluation of applied weaponry is done for each conflict measure. The highest value reached in a region within a month depicts the crucial value used to determine the intensity concerning the region and month.
Figure 2: Point System for the Indicator Use of Weapons
184.108.40.206. Indicator Use of Personnel
The quantity of both involved and deployed groups of people determines the assessment of personnel. In the case of state actors, these groups of people normally comprise soldiers and policemen, while in the case of non-state actors, they comprise armed followers and mercenaries. The number of personnel looked at and crucial for the evaluation counts those individuals acting as a collective group and representing one of the involved conflict actors within the framework of a violent form of communication. Hence, peaceful protesters may be involved in a violent form of communication as addressees, and are therefore considered relevant deployed personnel.
If the number of deployed personnel comprises up to 50 people, it is considered low. If the number of deployed personnel comprises up to 400 people, it is considered medium. If the number of deployed personnel exceeds a total of 400 people, it is considered high.
The evaluation of involved personnel is done for each conflict measure. The highest value reached in a region within a month depicts the crucial value used to determine the intensity concerning the region and month.
Figure 3: Point System for the Indicator Use of Personnel
2.2.2. Consequences of an Act of Violence – Casualties, Refugees and Destruction
Regarding the consequences of an act of violence, both the number of casualties directly linked to a violent conflict measure and the threat to livelihood of the population that is directly affected by the conflict measure, are taken into account. The latter is assessed by measuring the number of refugees and the degree of destruction.
220.127.116.11 Indicator Casualties
If the number of casualties within a particular region and month lies between 0 and 20, it is considered low. If the number of casualties lies between 20 and 60, it is considered medium. If the number of casualties exceeds 60, it is considered high. Evaluating the total number of casualties follows the principle of assessing the number of individuals who died as a direct consequence of a conflict measure within a certain month and region.
Figure 4: Point System for the Indicator Casualties
18.104.22.168 Indicator Refugees
For determining the level of intensity, both individuals fleeing across state borders (refugees) and internally displaced persons (IDP) are taken into account. However, only those people who flee as a direct consequence of a conflict measure are considered. Those people who had a refugee or IDP status previously to the conflict measure are not included in the calculation.
If the number of refugees is between 0 and 1,000 in a certain region and month, it is considered low. If the number of refugees is between 1,000 and 20,000 in a certain region and month, it is considered medium. If the number of refugees exceeds 20,000, it is considered high.
Evaluating the total number of refugees for the region within or from which people flee follows the principle of assessing the sum of individuals fleeing as a consequence of all conflict measures in a month and within or from said region.
Figure 5: Point System for the Indicator Refugees
22.214.171.124 Indicator Destruction
Destruction can be subdivided into four categories: (a) infrastructure (civilian and military), (b) habitation, (c) economy and self-sufficiency, and (d) identity-establishing goods.
It matters whether the observable destruction is massive in its scale. The destruction in one of the aforementioned categories is massive, if the damaged object was crucial for the functioning of the category in question. The level of destruction is low if either none occurred or none of the four aforementioned categories has experienced massive destruction. If massive destruction occurred in one or two of the categories, the destruction is considered medium. Destruction is considered high, if massive destruction occurred in three or four categories.
To evaluate the scale of the destruction, the sum of all destruction which occurred as a consequence of conflict measures carried out in a certain month and region is taken into account.
Figure 6: Point System for the Indicator Destruction
2.2.3 Evaluating the Indicators
The evaluation of the indicators Use of Weapons and Destruction is done qualitatively, that is by assessing the sort of a weapon, the use of said weapon or the destruction of a relevant object and by taking into account the context.
The evaluation of the indicators Personnel, Casualties, and Refugees is generally done quantitatively, that is by taking into account the listed thresholds. This approach, however, is only applicable if there is reliable data. (In this respect, the listed thresholds are rather guidelines.) If there is no reliable data, a qualitative evaluation is done by interpreting texts, especially news. The interpretation of texts becomes a systematic approach by applying a scheme of keywords which are assigned to the indicators. Table 8 shows examples of possible keywords which can be used to evaluate an intensity.
|Personnel||fire/maneuver team, squad/crew, several, patrol, platoon||company, bataillon, many, large number||regiment/brigade, division, corps, bataillon, large number|
|Casualties||several, dozen||many, dozens of, scores of, numerous, huge/high/large numbers||hundreds of, thousands of, several thousands of, huge/high/large numbers|
|Refugees||small number, several, dozen, dozens of, numerous, hundreds of||thousands of, several thousands of, huge/high/large numbers||ten thousands, huge/high/large numbers|
Table 2: Keywords for a Qualitative Evaluation
If an evaluation is not even possible through text interpretation, the HIIK methodology allows the conflict researcher to interpret the information based on his or her knowledge and expertise of the conflict and his or her experience. Taken together, the approaches form a traffic lights system of reliability to determine the intensities, as visualized in table 3.
|Quantitative determination||Reliable data|
|Qualitative determination||Text interpretation, keyword analysis|
|Informed interpretation||Knowledge of the conflict|
Table 3: „Traffic Lights of Reliability“
2.3 Determining Conflict Intensities
As mentioned in 2.1, the primary unit of analysis for conflict intensities is temporarily a calendar month, and geographically a region, that is the top-level, subnational, political division of the state. Only intensities with the levels 3, 4 or 5, that is violent conflict intensities, are determined on the level of a region and month. If a conflict is non-violent in a month and region, there is no determination of the intensity for this month and region.
Apart from the primary unit of analysis, there is a secondary unit of analysis on a higher level of aggregation. The secondary unit of analysis is temporarily a calendar year, ad geographically the conflict region. On the aggregated level of a year and conflict region, intensity levels 1 and 2 are determined, additionally to intensity levels 3, 4, and 5. The conflict region comprises all regions where a political conflict was fought in a certain year. If no violence occurred in one of the relevant conflict regions, the entire country, on whose territory the conflict is fought, is considered the conflict region.
Table 4 visualizes which intensities are determined on the different units of analysis.
|Subnational conflict region||Aggregated conflict region|
|Calendar month||levels 3, 4, 5|
|Calendar year||levels 1, 2; 3, 4 und 5|
Table 4: Primary and Secondary Units of Analysis and Conflict Intensities
The primary unit of analysis, month and region, is aggregated to a regional monthly intensity by combining the individually evaluated indicators listed in 2.2. and presented as a scheme in figure 7. Table 4 provides an overview of the calculation of a combined intensity, which is done by considering the individual points for each indicator.
Figure 7: Scheme of the Consolidation of all Individual Indicators into one Conflict Intensity
|Means of a violent measure||Weapons|
|0 points||1 point||2 points|
|Personnel||0 points||0 points||0 points||1 point|
|1 point||0 points||1 point||2 points|
|2 points||1 point||2 points||2 points|
|0 points||1 point||2 points|
|Refugees||0 points||0 points||0 points||1 point|
|1 point||0 points||1 point||2 points|
|2 points||1 point||2 points||2 points|
|Consequences of violence||Casualties|
|0 points||1 point||2 points|
|Existential threat||0 points||0 points||0 points||1 point|
|1 point||0 points||1 point||2 points|
|2 points||1 point||2 points||2 points|
|Regional monthly intensity||Means|
|0 points||1 point||2 points|
|Consequences||0 points||level 3||level 3||level 4|
|1 point||level 3||level 4||level 5|
|2 points||level 4||level 5||level 5|
Table 5: Calculation with all Individual Indicators to Establish the Intensity of a Violent Conflict
Accordingly, on the level conflict region/year proceed as follows:
If violence occurred in at least one region in at least one month, then the respective highest regional monthly intensity determines the conflict region yearly intensity.
This intensity has to be upgraded or downgraded depending on the indicators casualties and refugees:
- Upgrade from 3 to 4, if more than 360 (18 times 60) casualties or more than 18,000 (18 times 1,000) refugees occurred in the whole conflict region in the course of the entire year.
- Upgrade from 4 to 5, if more than 1,080 (18 times 60) casualties or more than 360,000 (18 times 20,000) refugees occurred in the whole conflict region in the course of the entire year.
- Degrading from 4 to 3, if less than 120 (6 times 20) casualties and less than 6,000 (6 times 1,000) refugees occurred in the whole conflict region in the course of the entire year.
- Degrading from 5 to 4, if less than 360 (6 times 60) casualties and less than 120,000 (6 times 20,000) refugees occurred in the whole conflict region in the course of the entire year.
It has to be determined if the intensity level is a 1 or a 2 (as illustrated in 2.1), if no violence occurred in at least one month in at least one region.
 Wright, Quincy (1942): A Study of War. Chicago: The University of Chicago Press; Richardson, Lewis Fry (1960): Statistics of Deadly Quarrels. Pittsburgh: Boxwood Press.
 Singer, David J.; Small, Melvin (1972): The Wages of War 1816-1965. A Statistical Handbook. New York, London, Sydney, Toronto: John Wiley & Sons.
 Gledditsch, Nils Petter; Wallensteen, Peter; Eriksson, Mikael; Sollenberg, Margareta und Strand, Havard (2002): Armed Conflicts 1946 – 2001: A New Dataset, in: Journal of Peace Research 39 (5): 615-637.
 Kende, Istvan (1971): Twenty-Five Years of Local Wars, in: Journal of Peace Research 8 (1): 5-22;
ebd. (1972): Local Wars in Asia, Africa and Latin America 1945-1969. Budapest: Center for Afro-Asian Research; ebd. (1982): Kriege nach 1945. Eine empirische Untersuchung. Frankfurt a. Main: Haag und Herchen.