Some thoughts on the local elections in Bosnia

A pattern seems to be emerging in the Central Bosnia canton, one of the two “mixed” cantons in the Bosniak-Croat Federation that was established in 1994.


In the 2004 mayoral elections, Jajce, previously a Croat-controlled municipality having been captured from the Serbs by the HVO (Bosnian Croat Army) in the dying days of the Bosnian war, elected an SDA (Bosniak nationalist) mayor. In the 2008 mayoral elections, the SDA took Busovaca and Novi Travnik, previously part of a crescent-shaped sliver of Croat-dominated municipalities surrounded by Bosniak-majority territory.

Yesterday it was the turn of Vitez, which became the fourth previously Croat-controlled municipality in the canton to come under the control of the SDA, although an examination of the voting figures suggests one should not read too much into this particular result. The percentage of the electorate voting for Croat mayoral candidates in Vitez actually rose from 60% in 2008 to 62% in 2012, with the SDA profiting from a split in the Croat vote. The Croat vote also held up at 46% in Busovaca, which was a straight contest between the Croat nationalist HDZ and the victorious SDA, and in Jajce, the Croat vote rose to 46% from 44% in 2008.

But in Novi Travnik, which the SDA held, the Croat vote fell from 56% in 2008 (when the SDA benefited from a split in the Croat vote) to 38% in 2012, while in Kiseljak, which the HDZ won, it declined to 64% from 70%.  In Zepce, a Croat enclave in the Bosniak-dominated Zenica-Doboj canton, 55% of votes went to explicitly Croat parties, compared with 60% last time, although the left-wing SDP, which is seen by many Croats as a party representing Bosniak interests, fielded a Croat candidate rather than a Bosniak one, which may have swung some Croat voters.

The picture is complex, but the dramatic falls in the Croat nationalist vote as a percentage of the total in Novi Travnik, Kiseljak and Zepce point to a weakening of the Croat position in the Central Bosnia and Zenica-Doboj cantons, a trend that is symbolised by the SDA’s victory in Vitez yesterday.

Local elections and ethnic consolidation in Bosnia

It is interesting to note that in the Muslim-Croat war of 1993-1994, neither side was able to capture any municipalities in which the other group had been in the majority on the eve of the war.

The most bitter fighting was over the municipalities in which no group had been in the majority according to the 1991 census, such as Mostar, Bugojno, Zepce and Fojnica. Most such municipalities were by the end of the fighting either like Mostar territorially divided between Croats and Muslims (known as Bosniaks from 1994) or under the control of the group that had made up the largest percentage of its population. The exceptions were two municipalities that prior to the war had a Muslim plurality but fell under Croat control, Zepce and Stolac, and another, Vares, which previously had a Croat plurality but was captured by the Bosnian Army (ARBiH) in October 1993.

In 1995, the Bosnian Croat Army (HVO) captured Jajce, another municipality that prior to the war had a Muslim plurality, from the Serbs,so at the end of the Bosnian war, the Croats controlled three municipalities where Bosniaks had previously been the largest group, while the Bosniaks controlled one where Croats had been most numerous.

Jajce became part of the Central Bosnian canton, one of the two “mixed” cantons that had been established by the Washington Agreement of 1994 that ended the Muslim-Croat war, while Stolac became part of the other “mixed” canton, Herzegovina-Neretva. Although “mixed”, it was quite clear when these cantons were established that the Bosniaks would be stronger in the Central Bosnia canton, while Croats would be the predominant group in Herzegovina-Neretva. Many Bosniaks returned to Jajce, often in the face of opposition from local Croats, and the municipality in 2004 elected a Bosniak mayor, but Stolac remains under firm Croat control.

Jajce is not the only municipality in the Central Bosnia Canton to move from Croat to Bosniak control. In the 2008 local elections, mayoral candidates from the main Bosniak political party, the SDA, won in Novi Travnik and Busovaca, which previously had Croat mayors. This is interesting because prior to the war they both had – albeit marginal – Croat pluralities. Novi Travnik and Busovaca may come to be seen as the second and third previously Croat-plurality municipalities to “fall” to the Bosniaks, 14 years after  Vares. The Croat “loss” of these three municipalities means that the Bosniaks have a 3-2 advantage in terms of control of municipalities where the other group was more populous in 1991. Elsewhere in the central Bosnia canton, Novi Travnik and Busovaca could even eventually be joined by Vitez, the previously Croat-plurality municipality that connects them and possibly even Kiseljak which was 52% Croat and 41% Muslim according to the 1991 census, so would be the first municipality in which one group had been in the majority to come under the control of the other group.

Zepce, which is in the Bosniak-dominated Zenica-Doboj canton, still has a Croat mayor, but this is something of an anomaly, because its borders were altered in 2001 to take in mainly Croat areas of the neighbouring municipalities of Zavidovici and Maglaj. Nevertheless the border alterations may not be enough to secure Croat predominance there permanently.

The SDA is highly unlikely to take control of Zepce, Vitez and Kiseljak in the local elections next month and Dobretici in the central Bosnia canton and Usora in Zenica-Doboj, two tiny Croat municipalities that were created after the Dayton accords, will elect Croat mayors as will Kresevo in the Central Bosnia canton. But the election results are likely to show a gradual consolidation of Bosniak predominance in these cantons.

In the Hercegovina-Neretva Canton, meanwhile, Stolac is likely to remain under firm Croat control despite the previous Muslim plurality and substantial Bosniak returns to that municipality. Interestingly, and perhaps in recognition of which way the wind is blowing, the SDA will not field a mayoral candidate there, or in Capljina, another municipality in the canton that before the war had a large Croat population.

The immediate fallout of the Muslim-Croat war was a rigid but very untidy division between the two populations, with areas controlled by each group spattered across territory dominated by other group. Conversely, Croats and Bosniaks are now much more intermingled, but it seems that a neat division corresponding to the Croat- and Bosniak-dominated cantons, including the "mixed" cantons, is emerging. Next month’s local elections could show this trend continuing.

With the benefit of hindsight: Croatian operations in Bosnia in 1995

Croatian Army (HVO) and Bosnian Croat Army (HVO) operations in Bosnia in 1995 helped pave the way for the much-celebrated defeat of the Republika Srpska Krajina (RSK), the part of Croatia that had been occupied by Serbs since 1991. Other benefits of these operations, which continued until well after the RSK collapse, are less easy to discern.

The first major success of the HV and HVO in Bosnia in 1995 was the capture of Bosansko Grahovo,a town in Western Bosnia, in July. This put the Croats in a position to advance towards Knin, the capital of the RSK, from two directions, which they did the following month during “Operation Storm”, the action that led to the collapse of the RSK. The operation in July also saw the Croats capture Glamoc, a town to the West of Bosansko Grahovo. Though part of an operation that was directed at Knin, the capture of Glamoc also put the Croats in a strong position to advance towards Jajce, from which they had been expelled following the Serb takeover in 1992. Bosansko Grahovo and Glamoc were prior to the war majority Serb towns with negligible Croat populations, so were of no interest from an ethnic point of view, but Jajce had been 35% Croat and was an attractive target.

The Serbs were on the ropes after Operation Storm and the HV/HVO and the Bosnian Army (ARBiH) began advancing in September, the HV/HVO towards Jajce and the ARBIH south and east out of the Bihac pocket, which had been besieged by the Serbs before Operation Storm, and in parallel with the northward Croat advance towards Jajce.

The Croats captured Drvar, Sipovo and Jajce, while the ARBiH’s Bihac-based 5^th Corps captured Sanski Most, Bosanski Petrovac, Kljuc and Bosanska Krupa and its 7^th Corps took Donji Vakuf. The ultimate prize for both the HV/HVO and the ARBiH was Banja Luka, the biggest city in Serb-controlled Bosnia. Capturing Jajce put the Croats in a position of strength vis-a-vis the ARBiH and prevented a link up between the 5^th and 7^th Corps. Nevertheless, the 5^th Corps was still in a possible position to advance on Banja Luka and, following its successes earlier in September, advanced towards Mrkonjic Grad. Further north it was advancing on Bosanski Novi and Prijedor.

Keen to advance towards Banja Luka before the ARBiH, the HV on 18 September opened up a new front, attacking the Bosnian Serb Army (VRS) across the river Una at three points along the Bosnia-Croatia border between Bosanski Novi (described Novi Grad on the above map) and Bosanska Dubica. The attack was a total failure, resulting in many HV deaths and a retreat back across the river.

The Croats made a final push towards Banja Luka in October, capturing Mrkonjic Grad, but under strong American pressure failed to advance any further.

The Croats captured six Bosnian towns during 1995. They also pushed the Serbs away from the Western Bosnian border with Croatia, which must have seemed like an advantage at the time, and gained control of Bosnian territory separating the Bosniak (Bosnian Muslim) heartland around the cities of Sarajevo, Zenica and Tuzla and the Bosniak-dominated Bihac pocket, which could be an advantage in any future negotiations with the Bosniaks. Eastern Slavonia, the part of Croatia bordering Serbia, was still under Serb control at this time, so capturing Serb territory in Bosnia may have helped to persuade the Serbs that they would have to give it up.

But Croat gains seem meagre with the benefit of hindsight. Five of the six towns they captured, Bosansko Grahovo, Glamoc, Drvar, Sipovo and Mrkonjic Grad, were sparsely populated and had previously had large Serb majorities. The latter two were returned to Serb control by the Dayton agreements.

And the other town they captured, Jajce, had actually been slightly more Muslim (39%) than Croat (35%) before the war. The Washington Agreement that ended the conflict between the Croats and the Muslims in 1994, establishing a Muslim-Croat Federation made up of cantons,  placed Jajce in a “mixed” canton in which Bosniaks were more numerous than the Croats rather than in a Croat-majority canton. Though the HVO capture of Jajce ensured that it was in reality under Croat control in the years after the war, facilitating the return of its Croat population, Bosniaks have also been returning and are now more numerous. The contours of a division between Croat and Bosniak spheres based on cantonal borders within the Muslim-Croat Federation may gradually be taking shape. The Bosnian newspaper Dnevni List earlier this year reported that the Federation’s Prime Minister Nermin Niksic highlighted the financial benefits of merging some of the cantons. Such a reorganisation would merge the “mixed” central Bosnian canton, which includes Jajce, with other Muslim-majority cantons, while the other “mixed” canton (centred on the city of Mostar and in which Croats are more numerous than Bosniaks) would merge with two Croat-majority cantons. So even the symbolically important capture of Jajce may not turn out to be a lasting victory from a Croat point of view.

Aside from the questionable territorial gains the Croats made in Bosnia in 1995, they also seem to have been motivated by the desire to curry favour with the West, which wanted to bring the Serbs to the negotiating table and to confine them to 49% of Bosnia’s territory, and to gain the status of a regional power. Both objectives were met to an extent, although the claim to have achieved the latter looks rather empty in retrospect, particularly in light of the demise of the planned confederation between Croatia and the Federation, which was still on the agenda before the Dayton agreement. To actually gain enough power to redraw the Bosnian map in their favour, the Croats would have had to capture Banja Luka, giving them strong leverage over both Serbs and Bosniaks. But they failed to do this and now see themselves as an oppressed minority in a Muslim-dominated Federation. Many must wonder whether it was worth engaging in military operations after Operation Storm at all.

Serb villages "under the protection of the Bosnian state"

“The Bosnian government is not a mirror image of Karadzic's regime: the mass murder of civilians is not one of its military objectives. Serb villages in reconquered areas of Herzegovina live peacefully now under the protection of the Bosnian state.”

This claim appears in an article by the historians Mark Almond, Adrian Hastings, Branka Magas, Norman Stone and Noel Malcolm published in the International Herald Tribune on 29 November 1994, two months after the Bosnian Army (ARBiH) captured about 100 square kilometres of territory south of the town of Konjic in September 1994.

The focus of the attack was the Bijela, a village that prior to the war had 635 Serb, 1,186 Muslim and 1,492 Croat inhabitants. The Committee for Collecting Data on Crimes Committed Against Humanity and International Law in Belgrade claimed that when the ARBiH took the village on 12 September they found only three people, “a bed-ridden old man and two mental patients”, killed the old man and transferred the other two to a prison camp in Konjic. The name of the old man (Simo Nenadic, born 1910) appears on a large list of names compiled by the Committee of Serbs allegedly killed in Konjic between 1992 and 1995. While this list should obviously be treated with scepticism, many of the same names appear on the widely-respected Sarajevo-based Research and Documentation Centre’s list of people killed in the war in Konjic, though Simo Nenadic’s does not.

The Committee’s implicit claim that the Serb inhabitants of Bijela left before the ARBiH advance is easier to establish. Speaking during a session of the National Assembly of the Serb People in Bosnia Herzegovina in Pale in November 1994, General Zdravko Tolimir, Assistant Commander for Intelligence and Security of the Bosnian Serb Army Main Staff,  said: “During the attack on Borci [a Serb-held town in Konjic municipality that was also targeted by the ARBiH in September] we have lost the villages of Bijela, Mladeskovici, Ljubina ... The total number of evacuated Serbian inhabitants is 850.”

According to an International Crisis Group  report of 1998, there were 728 Serbs in Konjic municipality in 1998. The report says that 600 Serbs left the municipality at the end of the war, suggesting a population of about 1,300 Serbs at the end of hostilities. Most of these Serbs probably lived in areas that were under Serb control before being transferred to the Muslim-Croat Federation in line with the Dayton Accords that ended the war. Given that there were no major changes in territorial control between the ARBiH advance in 1994 and the Dayton Accords, it seems reasonable to surmise that the Serb population in the entire territory Konjic municipality was roughly 1,300 in late 1994. If this number and Zdravko Tolimir's estimate of 850 evacuated Serbs are remotely correct, there was clearly a sharp fall in and possibly complete disappearance of the Serb population in the captured territory.

That these villages were unlikely to come “under the protection of the Bosnian state” in any meaningful sense is certainly suggested by previous actions of the ARBiH in Konjic municipality. When Muslim and Croat forces captured the village of Bradina in May 1992, inhabitants were taken to the Celebici prison camp and beaten, raped and murdered. Units of the ARBiH were then involved in atrocities against Croat civilians in Konjic during 1993, including the Trusina massacre in April of that year. During 1993, military units including the “Black Swans” and the 4th Muslim Brigade of the ARBiH’s Mostar-based 4^th Corps, or “Muderis”, commanded by the cleric Nezim Halilovic, gained notoriety in Konjic.

The 4^th Muslim Brigade  “played a key role” in the successful 1994 attack on Bijela and other villages, according to the journalist Sefko Hodzic in his book ‘Bosnian Warriors’. Elsewhere in Bosnia, in October 1994, the month before the publication of the Herald Tribune article, some 2,000 Serb civilians fled before the ARBiH 7^th Corps advance towards Kupres, according to the UN.

With the benefit of hindsight, we know that in 1995 the much more extensive ARBiH capture of territory was preceded by the flight of the Serb inhabitants, wisely given the treatment of those who were unable to leave. In northern Bosnia, the majority Serb village of Vozuca was taken by the ARBiH in September 1995, to provide a road link between its 2^nd and 3^rd Corps, but also, according to some sources, to house refugees from Srebrenica, which had recently fallen to the Serbs. The attack, assisted by foreign “Mujahideen” fighters, resulted in war crimes and the exodus of Vozuca's Serbs.

The actions of the ARBiH obviously have to be seen in the context of the much worse atrocities committed by the Serbs, but the idea that the Serb villages it captured came “under the protection of the Bosnian state” – and the implication that Serb civilians remained there and were left unharassed – is not credible.

Ethnic cleansing in Bijeljina

Bijeljina is an interesting case study because it was one of a handful of municipalities in Bosnia that, except for a few days of fighting at the beginning of the war, the Serbs took over without opposition.

Examining what happened there during the war sheds light on the long-term political aims, as distinct from short-term military aims, of the Bosnian Serb leadership, because, unlike in other areas, the ethnic cleansing visited on the Muslim population cannot simply be dismissed as a side-effect of the fighting.

And Bijeljina was by the end of the war in 1995 thoroughly ethnically cleansed of Bosnian Muslims – much of this cleansing having been accomplished in 1992 – suggesting that an ethnically pure Republika Srpska (RS) was indeed the political aim of the Bosnian Serb leadership.

The ethnic cleansing was facilitated by killing, some of it during the takeover of the town when Serb paramilitaries did face opposition from the Muslim ‘Patriotic League’, but most of it when the town had already been secured and the Muslim population could not meaningfully be perceived as threatening Serb control of the town. A report by the influential Balkan Investigative Reporting Network, ‘Bijeljina’s Strange Silence over War Crimes’, notes that the Sarajevo-based Research and Documentation Centre found that 1,040 people were killed in Bijeljina during the war. The report fails to note that a large majority of this figure was actually Serb, presumably soldiers from the municipality killed on the nearby frontline, but still, some 300 Muslims were killed, which at about 1% of the municipality’s pre-war population is a substantial figure, probably enough to scare the rest of the population into leaving.

The ethnic cleansing of Bijeljina’s Muslims is described in the 2005 International Criminal Tribunal for the Former Yugoslavia (ICTY) judgement against Momcilo Krajisnik, a senior member of the Bosnian Serb leadership.

According to this judgement, Serb paramilitaries, together with local MUP (Interior Ministry) forces, terrorised Muslims and some Serbs in Bijeljina in the months following the takeover. On 15 June the paramilitary leader Ljubisa Savic, known as Mauzer, stated that the presidency of SAO Semberija-Majevica (the region including Bijeljina and neighbouring municipalities) had decided that all Muslims in managerial positions would be fired if the “genocide” continued against Serbs elsewhere in the country. From July, Muslims in the town were subjected to looting and expulsion by the SDS. In this, Vojislav “Vojkan” Durkovic of the Bijeljina SDS was aided by Mauzer’s men. Muslim houses were then allocated to Serbs in exchange for a fee. “The Bijeljina SDS was determined to rid the municipality of its remaining Muslims”by killing a Muslim family on each side of town. The plan was implemented in September 1992 by a special police unit led by Dusko Malovic at the instigation of Drago Vukovic of the MUP. The Serb plan to rid Bijeljina of its Muslims also involved cutting off electricity, water and telephone lines to the homes of Muslims who refused to be mobilised and firing them, as well as forcing prominent Muslims to perform menial tasks. At least 52 people, mostly Muslims, were killed between April and September, on top of the at least 48 civilians killed during the takeover of the town.

This account of events in Bijeljina during 1992 is almost wholly taken from the witness statement and testimony of Milorad Davidovic, a policeman working for the Yugoslav Federal SUP (Secretariat for Internal Affairs) who was called in by the SDS leadership in July 1992 to help deal with problems caused by paramilitaries in Bijeljina, but soon forced out, probably because he was seen as being to rigorous.

The description of Djurkovic as “of the SDS” comes from Davidovic’s account. According to Davidovic’s witness statement Djurkovic “worked for the SDS as a field operative.” But the Krajisnik judgement does not explain Djurkovic’s role in the SDS in Bijeljina and fails to note that he was actually a member of the Party of Serbian Unity, which was headed by Arkan, the paramilitary leader active during the takeover of Bijeljina in early April 1992, setting in train the reign of terror continued by Djurkovic and Savic. Savic was a member of the SDS, but the relationship of both of these characters with the Bijeljina SDS leadership and the SDS leadership for the whole of RS is not fully explained.

The Krajisnik judgement also draws heavily from Davidovic’s account of a three-part SDS plan to ethnically cleanse Bijeljina of its Muslims. The plan involved: killing Muslim families in different parts of the city to scare the remaining Muslims away; firing Muslims who refused to respond to the call for mobilisation from their jobs and cutting off their utilities; and forcing prominent Muslims to perform menial tasks. The first and second parts of the plan were conducted by Drago Vukovic and his colleague Predrag Jesuric, while the third part “was the SDS policy,” thought up by a member of the SDS municipal main board. The link between the SDS and the first two parts of the plan is not explained, although Davidovic does say that he attended a meeting of the local SDS and crisis staff where the whole plan was discussed. Predrag Jesuric was according to Davidovic, “the main ideologue in the Bijeljina SDS”, but later in his witness statement Davidovic says, “I do not know if he was a member of the SDS.” Drago Vukovic’s connection with the SDS is not explained.

Those inclined to swallow Davidovic's account of events in Bijeljina whole should note that he  has described Mirko Blagojevic, a local Serb Radical Party and paramilitary leader who is named as one of members of the Joint Criminal Enterprise in the Krajisnik judgement as a “very positive character during the war in terms of protecting Muslims”. He also stated that the conflict in Bijeljina was “started by the Muslim forces that tried to provoke a clash.”

Nevertheless, Davidovic’s account is valuable and it is still the case that the SDS was the main political force in Bijeljina and presided over the ethnic cleansing of Muslims in 1992. There is little reason to doubt Davidovic’s claim that Milan Novakovic, an SDS deputy from Bijeljina elected in 1990, was close to the paramilitaries operating in Bijeljina, aware of their activities and unwilling to do anything about them. (Davidovic says that another elected SDS deputy from Bijeljina, Dragoljub Micic, was sympathetic to his attempts to halt the ethnic cleansing activities of the paramilitaries in Bijeljina.) His claim that members of the Bosnian SDS leadership such as Radovan Karadzic and Momcilo Krajisnik benefited from and knew about the looting activities of the Serb paramilitaries in Bijeljina is also very credible.

That the SDS leadership was heavily implicated in but not in full control of the ethnic cleansing of Bijeljina is also suggested by an Amnesty International Report from 1994, which describes Vojkan Djurkovic’s“Commission for the Exchange of Population”, established in 1992 to conduct ethnic cleansing operations, as “semi-official”, but later vying for power with other Serb factions in the area.

The Amnesty report also cites the claim by the human rights group the Humanitarian Law Fund that in 1993 the Bijeljina authorities said that they were implementing a policy of to reduce the municipality’s Muslim population to 5%. There is no smoking gun evidence of this kind in the ICTY account of events in Bijeljina in 1992, however.

The appeal judgement in the Krajisnik case adds to the impression that events in Bijeljina in 1992 may not have been the result of a plan to drive out the entire Muslim population. It notes that since each of the Serb crisis staffs included at least one Assembly deputy among its members the Bosnian Serb leadership exercised “a substantial amount of control over” the crisis staffs, but also that the initial judgement against Krajisnik “did not reach any general finding on the link between the Bosnian-Serb leadership and crisis staffs.”

Many far-fetched claims have been made about the wars in the former Yugoslavia, by all sides. Among these are the Serb claim that the exodus of Serbs from the Croatian Krajina region was part of a Croat plan for an ethnically pure Croatia and the Croat claim (outlined in Charles Shrader’s “The Muslim-Croat Civil War in Central Bosnia") that there was a Muslim plan to drive the Croat population out of central Bosnia. The claim that what happened in Bijeljina in 1992 was wholly the result of a plan to create an ethnically pure “Greater Serbia” also belongs in this category.