![[PDF]](/pb-assets/Icons/adobe-pdf-icon-1498649896243.png){kind=icon}
Abstract
The dramatic increase in the number of fatalities in organized violence, seen between 2011 and 2014, did not continue in 2015 and 2016. Rather, the notation of some 131,000 fatalities in 2014 was followed by a steep decline, with just below 119,000 in 2015 and a little over 102,000 fatalities in 2016. Despite the decrease, the number was the fifth highest during the entire 1989–2016 period. Most of the fatalities – over 87,000 – were incurred in state-based conflicts, the main driver behind the trend. Just as the number of fatalities, the number of state-based conflicts, albeit remaining on a high level, continued to decrease in 2016, going from 52 to 49, with 12 of them reaching the level of war, with at least 1,000 battle-related deaths. Also the non-state conflicts dropped in number in 2016, from 73 to 60. This was followed by a decrease in the number of fatalities, and only one conflict caused more than 1,000 deaths. Twenty-one actors were registered in one-sided violence, down by five from 2015. A number this low has only been recorded twice before; in both 2009 and 2010, 21 one-sided actors were listed in UCDP data. The number of fatalities also decreased, going from almost 9,800 to a little over 6,000.
Introduction
This is the second year that the annual update from the Uppsala Conflict Data Program (UCDP) presents trends in not only state-based armed conflict, but also non-state conflict and one-sided violence.1 The three categories are coded as mutually exclusive and can be aggregated as ‘organized violence’.2
Organized violence 1989–2016
The three subcategories of organized violence share the same intensity cut-off – 25 fatalities in a calendar year.3State-based armed conflict includes all cases where at least one of the parties is the government of a state, that is, armed conflicts between states and within states (government vs. a rebel group). They include cases ranging from small-scale conflicts, only active briefly (e.g. the attempted coup in Lesotho in September 1998), to large-scale wars that rage for many decades (e.g. the government conflict in Afghanistan, active on a high-intensity level since 1978). Non-state conflicts include fighting between rebel groups, militias and drug cartels such as the conflict between the Islamic State (IS) and Taleban, raging since 2015 in Afghanistan. It also contains conflicts between informally organized groups, notably between groups with a common identification along ethnic, clan, religious, national, or tribal lines. Examples include interethnic violence between Birom and Fulani and between Fulani and Tiv in Nigeria’s Middle Belt. Finally, One-sided violence entails the targeted killing of unarmed civilians, by states (e.g. the Afghan Taleban government’s excessive killing of civilians, particularly in 1998, when it consolidated its control over the northern areas of the country), or formally organized non-state groups (e.g. the Ugandan rebel group LRA’s massacres across northern Uganda, eastern DR Congo, Central African Republic, and South Sudan).
Building on conflict theory, UCDP defines state-based armed conflict around political issues, that is, incompatibilities concerning government and/or territory that are disputed by force by two parties, of which at least one is the government of a state. (For full definitions, see the Online appendix.) Hence, state-based armed conflicts are political in this central aspect. In contrast, there is no requirement of any identifiable incompatibility in the definition of non-state conflict, which involves the use of armed force between two organized armed groups, neither of which is the government of a state. This means that what would typically be thought of as criminal violence, for example, turf wars between drug cartels, may show up in the non-state conflict category. However, much lethal criminal violence will never enter UCDP figures since the intensity threshold – at least 25 deaths between the same two organized groups in one calendar year – is not met, or because many of the criminal deaths are impossible to attribute to any identifiable actor. Thus, state-based armed conflict captures violent political conflict, whereas some criminal violence is coded as non-state conflict. Neither is there any requirement of any identifiable political issue in the definition of one-sided violence. Sometimes one-sided violence happens as part of criminal activities, for example the numerous targeted killings of civilians by the drug cartel Los Zetos in Mexico. Yet, both non-state conflict and one-sided violence often happen in close connection to wider political conflict.
Figure 1 reveals that the dramatic increase in the total number of fatalities seen between 2011 and 2014 did not continue in 2015 and 2016.4 Rather, the notation of some 131,0005 deaths in 2014 – second only to the death toll in 1994, the year of the Rwandan genocide – was followed by a steep decline, with just below 119,000 in 2015 and a little over 102,000 fatalities in
2016. However, despite this drop, the number remains one of the highest recorded in the entire time period.
State-based conflict 1946–2016
Since the end of World War II, 600 dyads have been active in 280 conflicts in 157 locations.6 Corresponding numbers for the post-Cold War period are 357 dyads in 171 conflicts in 93 locations.
There were 49 conflicts active in 2016,7 down by three from the previous year,8 which, together with 1991, was the peak year of the period. With this, the increase in the number of active conflicts, recorded since 2013, was broken. The steepest rise in the number of conflicts from one year to the next was between 2014 and 2015, from 42 to 52. This was driven almost completely by developments relating to the expansion of the Islamic State (IS) (Melander, Pettersson & Themnér, 2016). This expansion did not continue at the same rate in 2016. While two new countries were challenged by the Islamist group (Jordan and Pakistan), four of the
Of the 49 conflicts, two were fought between states: Eritrea–Ethiopia and India–Pakistan. As can be seen in Figure 2, the number of interstate conflicts has been low throughout the entire 1946–2016 period, and particularly so since the new millennium. Since the turn of the century, as many as seven years have passed with no conflicts fought between states, and the remaining ten years have seen either one or two active conflicts. While the India–Pakistan conflict has been active intermittently during these years, the one fought over Eritrea’s and Ethiopia’s contested common border was last active in 2000. Despite the signing of an accord in that year, the two states have failed to agree on the demarcation of the border and the relationship between them remains tense. In 2008, the United Nations Mission in Ethiopia and Eritrea was disbanded and since then there have been numerous rumors of border skirmishes, but these have rarely been possible to verify. In 2016, tensions heightened, and came to a head on 12 June, when a battle erupted, involving medium- and long-range artillery. While reliable information on the fighting was difficult to come by, this time both parties agreed that the fighting actually took place, with both sides accusing the other of initiating the two-day battle.
Of the 47 intrastate conflicts, as many as 18, or 38%, were internationalized, in the sense that external states contributed troops to one or both sides in the conflict. Both this number and share are very high, seen in the post 1946 context, second only to one other year: 2015, with 20 internationalized conflicts (39%). The proliferation of internationalized conflicts is worrying, since research has shown that external troop involvement tends to both make conflicts more lethal and prolong them (Balch-Lindsay, Enterline & Joyce, 2008; Lacina, 2006).
In 2016, 12 conflicts reached the intensity of war, with at least 1,000 battle-related deaths. An increase by one from 2015, this number is high seen over the most recent decade, but not in a longer perspective. As can be seen in Figure 3, the 1980s was the decade worst hit by
|
Table I. The ten most conflict-affected countries in terms of fatalities, 1989–2016
Looking beyond the number of wars to the number of battle-related deaths, Figure 1 reveals that the decrease seen in 2015 continued in 2016.9 With a death toll of almost 105,000 in the state-based category, the year 2014 was the most violent year in the entire post-Cold War period. That year, the peak of an unprecedented increase in fatality numbers was recorded (Pettersson & Wallensteen, 2015), driven mainly by an escalation of the conflicts in Syria, but also by increasingly heavy fighting in countries such as Iraq, Afghanistan, and Nigeria. Fatality numbers then dropped, to almost 98,000 in 2015, and then to a little over 87,000 in 2016.10 Despite this substantial decrease, the number remains at an exceptionally high level for the post-Cold War period, second only to the years 2013–15.
A little less than half of the fatalities recorded for state-based conflicts in 2016 were due to the conflicts in Syria, which caused over 40,500 fatalities, down by almost 5,000 from 2015. Thus, the overall decrease was mainly due to developments in Syria. UCDP records three state-based conflicts in the country: one over government power, one against IS, fighting for the establishment of an Islamic State, and one against the coalition SDF (Syrian Democratic Forces), which includes numerous groups, with goals ranging from ousting the government to control over Rojava, the Kurdish name for the northern region of Syria. The conflict over government has throughout been the deadliest, causing over 27,000 of the 40,500 deaths in 2016. It was followed by the conflict between the government and IS, with some 13,200 fatalities. The SDF conflict was by far the smallest, with around 150 battle-related deaths. The decline in the number of deaths from 2015 to 2016 was entirely driven by developments in the conflict over government power. The reason for the decline was mainly due to a reduction in fighting from 27 February and for some months, resulting from a ceasefire. The ceasefire was part of a negotiated deal, based on United Nations Security Council Resolution 2254, passed in December 2015. It did not include all active rebel groups, and attacks on IS and Jabhat Fateh al-Sham (commonly called al-Nusra) continued. While ravaged by difficulties from the start, the ceasefire did play a part in reducing the level of violence in March–May. However, despite this decrease in fatalities, the conflict was still by far the deadliest globally. As shown in Table I, Syria is the country with by far the largest aggregate number of casualties in state-based conflict after the end of the Cold War. Furthermore, Syria is only surpassed by Rwanda in total casualties.
Looking at fatality numbers broken down by region, Figure 4 depicts how the center of gravity has moved between regions over the time period. Africa saw the bulk of fatalities in state-based conflicts early in the period, as well as in the late 1990s, with both peaks mainly driven by developments on the Horn. In 1989–90, the conflict between the Ethiopian Derg regime and a range of rebel groups caused high death tolls, and in 1999–2000 the interstate war between Ethiopia and Eritrea was reminiscent of bloody World
War I trench warfare. The region also saw high fatality numbers in 1997–98, mainly driven by conflicts in DR Congo, Republic of Congo, and Sudan.
In the first decade of the 21st century, and especially the years 2006–11, the center of gravity shifted to Asia. In 2006–07 and onwards, the conflict in Afghanistan escalated after a period of less intense fighting, causing large numbers of deaths. Further driving the trend was the escalation of the conflict between the Pakistani Taleban, TTP (Tehrik-i-Taleban), and the Pakistani government from 2008 onwards, as well as the bloody end phase of the war in Sri Lanka, in 2008–09.
For most of the period, the Middle East was not one of the regions worst hit by state-based conflict. However, in 2003, the invasion of Iraq by the USA and its allies caused large numbers of deaths, and in 2012, the center of gravity shifted to the region again, and has remained there ever since. As discussed above, this is by and large due to the wars in Syria, which escalated steeply in 2012; the country has remained the deadliest worldwide since then in terms of battle-related deaths. It is worth noting that Syria is not the only country in the Middle East hit hard by conflict, however, as the war in Iraq has caused more than 10,000 fatalities in all years since 2014, when it escalated markedly as IS launched a large-scale offensive, quickly seizing large swathes of territory.
Non-state conflict 1989–2016
A total of 624 non-state conflicts have been recorded by UCDP since 1989, with a yearly average of 36 active conflicts. The upward trend in the number of active non-state conflicts, recorded since 2010, did not continue in 2016. In 2016, UCDP recorded 60 such conflicts,11 compared to 73 the previous year. While the number of non-state conflicts tends to oscillate from one year to the next, the change from 2015 to 2016 was one of the most substantial in the entire 1989–2016 period. Despite the considerable decrease, 2016 still had the third highest number of non-state conflicts for the period 1989–2016.
Looking at the geographical distribution of non-state conflicts, the majority of them took place in Africa (33 of the 60 recorded conflicts in 2016). Africa saw the same number of conflicts in 2016 and 2015, but the conflicts were not all the same. Whereas several of the 2015 non-state conflicts were not active in 2016, a number of new conflicts erupted in 2016 and several cases reignited. Some notable examples of the latter were the conflict between the ethnic groups Anuak and Nuer, in Ethiopia, and that between Marakwet and Pokot, in Kenya, neither of which had been active since the early 2000s. The Middle East experienced 17 non-state conflicts in 2016, a substantial decrease from the 23 recorded in 2015. This drop can partly be explained by fewer active non-state conflicts in Syria. The Americas had eight active non-state conflicts in 2016, a decrease from 12 in 2015. Since 2012, all non-state conflicts in the Americas have taken place in Mexico. Finally, two non-state conflicts were recorded in Asia in 2016, both in Afghanistan, a decrease by three from the previous year.
As in recent years, the most common type of non-state conflict in 2016 was between formally organized groups, such as, for example, rebel groups, militias, and drug cartels, most notably in Syria and Mexico. UCDP recorded 38 conflicts between such actors, compared to 22 between informally organized actors, such as ethnic or religious groups. All of the active non-state conflicts in the Middle East were between formally organized groups and all but three were fought in Syria.
Looking at the number of people killed in this type of violence, Figure 5 reveals that it was at its lowest in 2005, at just over 2,100, and peaked in 2014, at nearly 12,900. From the peak year, it decreased to almost 8,800 in 2016. Despite this decrease, 2016 was the fifth deadliest year in the 1989–2016 period. In 2016, only one non-state conflict caused more than 1,000 battle-related deaths: the one between IS and SDF in Syria.
Even though Africa had considerably more active non-state conflicts in 2016 than the Middle East, the number of fatalities in the latter was more than twice the number for the former, as seen in Figure 5. About 62% of the fatalities recorded in active non-state conflicts in 2016 occurred in the Middle East, whereas only 27% occurred in Africa.
One-sided violence 1989–2016
A total of 254 actors in one-sided violence have been recorded by UCDP since 1989, with a yearly average of 32 active actors. The number tends to vary from one year to the next, but figures for the last decade indicate an overall downward trend compared to the 1990s and the early 2000s; as many as 45 actors were recorded in 2002 and 2003. In 2016, 21 actors carried out one-sided violence,12 a decrease by five compared to 2015. A number this low has only been recorded twice before; in both 2009 and 2010, 21 one-sided actors were listed in UCDP data.
Moving on to the number of fatalities in this type of violence, UCDP estimates show that at least 6,000 civilians were targeted and killed by a state or a formally organized actor in 2016. The figure is considerably lower than in 2015, when nearly 9,800 deaths were recorded. As shown in Figure 6, the number of deaths in the one-sided violence category has gone down since 2014. The 2016 figure is further considerably lower than the high levels of one-sided violence seen in much of the 1990s, with the peak year of 1994 when hundreds of thousands of civilians were killed in the Rwandan genocide.
The reduction in recorded fatalities in one-sided violence from 2015 to 2016 can partly be explained by developments in Nigeria, where IS and Jama’atu Ahlis Sunna Lidda’awati wal-Jihad (commonly called Boko Haram)13 have been weakened. The August 2016 change in regional leadership, which deposed Abubakar Shekau, also seems to have led to a behavioral shift away from large-scale attacks against Muslim civilians, to a focus on Christian civilians (which only make up around 2% of the population in the worst hit areas) and the military.
A clear majority of the recorded one-sided violence in 2016 was carried out by non-state actors, as opposed to states, which has been the case in recent years (see Figure 6). Despite the substantial decrease in IS killings in Nigeria, the group was still responsible for more than half (61%) of the total number of civilians killed in one-sided violence. The killings were carried out in several countries in Asia, Europe, and the Middle East, with the majority taking place in Syria and Iraq.14 Looking at the period 1989–2016, it is worth noting that IS was the only non-state actor that was active in multiple regions during the same year (other groups, for example al-Qaida, have been active in more than one region, but not during the same year).
Conclusion
The total number of fatalities in organized violence declined in 2016, from just below 119,000 in 2015 to a little over 102,000. This means that the world has seen a decline of more than 20% in the number of fatalities in the two years following the latest peak in 2014, when the number of people killed was around 131,000. Despite this good news, the number in 2016 was the fifth highest during the entire 1989–2016 period, and we have no way of knowing if this two-year reduction will become a new downward trend. As in all previous years except 1994, when the Rwandan genocide took place, the trend in fatalities is driven mainly by state-based armed conflicts. Of the 47 intrastate conflicts active in 2016, as many as 18, or 38%, were internationalized, in the sense that external states contributed troops to one or both sides in the conflict.
Replication data
The complete datasets (UCDP/PRIO Armed Conflict Dataset, UCDP Dyadic Dataset, UCDP Battle-Related Deaths Dataset, UCDP Non-State Conflict Dataset, and UCDP One-Sided Violence Dataset) updated to 2016 are found at http://ucdp.uu.se/downloads/. Older versions of these datasets can also be found at this address (all datasets) and www.prio.no/cscw/armedconflict (the UCDP/PRIO Armed Conflict Dataset). The tables and figures in this article were created directly from the Excel sheets available at the UCDP web page. Detailed descriptions of the individual cases are found in the UCDP Conflict Encyclopedia at www.ucdp.uu.se. Replication data for this article can be found both at http://ucdp.uu.se/downloads/ and https://www.prio.org/jpr/datasets/.
Acknowledgements
Numerous colleagues in Uppsala have contributed to the data collection, notably Simon Pierre Boulanger Martel, Mihai Croicu, Emma Elfversson, Garoun Engström, Daniel Finnbogason, Helena Grusell, Stina Högbladh, Gabrielle Lövquist, Henrik Persson, Therese Pettersson, Margareta Sollenberg, and Samuel Taub. Also summer interns Sebastian Hellberg and Victoria Lecomte were of great help. A special thanks for comments on the first draft by Gabrielle Lövquist, Therése Pettersson and Anders Themnér. We are grateful to Martin Tegnander for assistance with designing Figures 1 and 6.
Notes
1
For full definitions of all key concepts, see the Online appendix.
2
The data for all three categories included in organized violence go back to 1989; for state-based armed conflict, they extend back to 1946.
3
It is worth noting that as soon as a dyad/actor has exceeded the 25-fatality cut-off, all violent events involving this dyad/actor are recorded and coded and made available in UCDP’s georeferenced event dataset, UCDP GED, and also for years when the violence did not cross the threshold.
4
The decline was also visible for violence involving al-Qaida, IS, and affiliates, which was discussed in Melander, Pettersson & Themnér (2016). An updated version of the figure presented in said article (Figure 5), displaying fatalities in organized violence and separating out al-Qaida, IS, and their affiliates, is available in the Online appendix.
5
Figures used in the text are UCDP’s best estimates. For a definition of different fatality estimates, see the Online appendix.
6
Last year we incorrectly reported 159 locations due to a faulty pivot table. UCDP counts incompatibilities concerning government, territory, or both. While a state can only experience one intrastate conflict over government in a given year, it can be a primary party in several armed conflicts over government and/or territory.
7
For a full list of state-based conflicts in 2016, see Appendix 1.
8
Last year, we reported 50 active conflicts for 2015. Based on new information we have added two conflicts for this year: between the Algerian government and Jund al-Khilafah; and between the Turkish government and IS. Tables 1–2, Figure 1, the UCDP Conflict Encyclopedia (www.ucdp.uu.se), the UCDP Dyadic Dataset, and the UCDP/PRIO Armed Conflict Dataset have been amended accordingly.
9
Battle-related deaths are those incurred in direct fighting between the parties to the conflict, in guerilla tactics and bombings, and also those civilians that were caught in the cross-fire. For more on this, see the Online appendix.
10
While the overall trend – driven by events in Syria – was that of a decrease, conflict intensity remained or even escalated in other war-torn countries. For example, in Afghanistan, the total number of fatalities in the two wars fought increased and, at over 17,900, reached a level not recorded before in the post-Cold War period. Also the war in Iraq remained at a very high level of intensity in 2016, with more than 11,500 fatalities.
11
For a full list of non-state conflicts in 2016, see Appendix 2.
12
For a full list of actors engaged in one-sided violence in 2016, see Appendix 3.
13
In March 2015, Boko Haram leader Abubakar Shekau pledged allegiance to the Islamic State, and, upon acceptance, the group became integrated into the latter. However, in August 2016, the IS central command announced a change in the leadership for its West Africa province, effectively deposing Shekau. With this, Shekau and a faction of the group re-formed Boko Haram, and for the last five months of the year, the two were both active in the region.
14
While the number of recorded IS incurred fatalities in one-sided violence in Iraq was high, it does not capture the many civilians killed in, for example, the battle over Mosul. These fatalities are instead recorded as part of the state-based category due to a UCDP coding rule that establishes that all deaths from cross-fire, shelling, etc. should be recorded as battle-related deaths. The same applies for Syria, where the bulk of civilian fatalities are coded in the state-based category for this exact reason, and only a fraction as one-sided.
References
|
Balch-Lindsay, Dylan, Enterline, Andrew J, Joyce, Kyle A (2008) Third-party intervention and the civil war process. Journal of Peace Research 45(3): 345–363. Google Scholar | SAGE Journals | ISI | |
|
Lacina, Bethany (2006) Explaining the severity of civil wars. Journal of Conflict Resolution 50(2): 276–289. Google Scholar | SAGE Journals | ISI | |
|
Melander, Erik, Pettersson, Therése, Themnér, Lotta (2016) Organized violence, 1989–2015. Journal of Peace Research 53(5): 727–742. Google Scholar | SAGE Journals | ISI | |
|
Mesterhazy, Attila (2016) The International Military Campaigns against Daesh. Report prepared for the Defence and Security Committee, NATO Parliamentary Assembly. Report no. 163 DSCTC 16 E rev. 1 fin. Google Scholar | |
|
Pettersson, Therése, Wallensteen, Peter (2015) Armed conflicts, 1946–2014. Journal of Peace Research 52(4): 536–550. Google Scholar | SAGE Journals | ISI |
Appendix 1. State-based armed conflicts active in 2016
This list includes all conflicts that exceeded the minimum threshold of 25 battle-related deaths in 2016 and fulfilled the other criteria for inclusion.1 The column Year shows the latest range of years in which the conflict has been active without interruption. The start year is found in parenthesis in the Incompatibility column, which indicates when the armed conflict reached 25 battle-related deaths for the first time. If a conflict has been inactive for more than ten years or if there has been a complete change in the opposition side, the start year refers to the onset of the latest phase of the conflict. For more complete information on the conflict- and dyad history, see (a) the UCDP/PRIO Armed Conflict Dataset and the UCDP Dyadic Dataset at http://ucdp.uu.se/downloads/ and (b) the Uppsala Conflict Data Program’s online conflict encyclopedia at www.ucdp.uu.se. The column ‘Intensity in 2016’ displays the aggregated conflict intensity in terms of the number of battle-related deaths. Thus, if more than one dyad is active in the conflict, the intensity column records their aggregated intensity. Three fatality estimates are given in the table: low, best and high.
|
Unclear cases of state-based armed conflict in 2016
Cases that have been completely rejected because they definitely do not meet the criteria of armed conflict are not included in the list below. For the conflicts listed here, the available information suggests the possibility of the cases meeting the criteria of armed conflicts, but there is insufficient information concerning at least one of the three components of the definition: (a) the number of deaths, (b) the identity or level of organization of a party or (c) the type of incompatibility. The unclear aspect may concern an entire conflict or a dyad in a conflict that is included above.
|
Appendix 2. Non-state conflicts active in 2016
This list includes all non-state conflicts that exceeded the minimum threshold of 25 deaths in 2016 and fulfilled the other criteria for inclusion.1 The column ‘Start year’ shows the first year when the non-state conflict caused at least 25 fatalities (since 1989). The column ‘Fatalities in 2016’ displays the number of people killed, in the low, best and high estimate.
|
Appendix 3. One-sided violence in 2016
This list includes all cases of one-sided violence that exceeded the minimum threshold of 25 fatalities in 2016 and fulfilled the other criteria for inclusion.1 The column ‘Start year’ shows the first year when one-sided violence caused at least 25 fatalities (since 1989). The column ‘Fatalities in 2016’ displays the number of civilians killed, in the low, best and high estimate.
|
MARIE ALLANSSON, b. 1984, MA in Peace and Conflict Research (Uppsala University, 2009); project leader and information officer, Uppsala Conflict Data Program, Department of Peace and Conflict Research: articles on conflict data published in SIPRI Yearbook in 2012 and 2013.
ERIK MELANDER, b. 1969, PhD in Peace and Conflict Research (Uppsala University, 1999); Professor at the Department of Peace and Conflict Research and Director of the Uppsala Conflict Data Program, Uppsala University; articles published in journals such as Journal of Conflict Resolution, International Studies Quarterly, Journal of Peace Research, and Journal of Gender Studies; forthcoming book The Peace Continuum: What It Is and How To Study It (coauthored with Christian Davenport & Patrick Regan, Oxford University Press).
LOTTA THEMNÉR (previously Harbom), b. 1975, MA in Peace and Conflict Research (Uppsala University, 2002); project leader, Uppsala Conflict Data Program, Department of Peace and Conflict Research: articles on conflict data published in the Journal of Peace Research and SIPRI Yearbook since 2005; editor of States in Armed Conflict between 2004 and 2012.
