The European Union Agency for Network and Information Security (ENISA), the European Defence Agency (EDA), the European Cybercrime Centre (EC3) and the Computer Emergency Response Team for the EU Institutions, Agencies and Bodies (CERT-EU) today signed a Memorandum of Understanding (MoU) to establish a cooperation framework between their organisations.
The Memorandum of Understanding was signed by Udo Helmbrecht, ENISA's Executive Director, Jorge Domecq, Chief Executive of the EDA, Steven Wilson, Head of EC3 and Ken Ducatel, CERT-EU's Acting Head.
The MoU aims at leveraging synergies between the four organisations, promoting cooperation on cyber security and cyber defence and is a testament to the trusted partnership that exists between these EU agencies. More specifically, it focuses on five areas of cooperation, namely Exchange of information; Education & Training; Cyber exercises; Technical cooperation; and Strategic and administrative matters. It also allows for cooperation in other areas identified as mutually important by the four organisations.
This collaboration will ensure the best possible use of existing resources by avoiding duplicative efforts and building on the complementarity of ENISA, EDA, EUROPOL and CERT-EU. This framework brings added value to the expertise, support and services that these parties provide to the European Union organisations, Member States and all stakeholders concerned.
High Representative/Vice-President and Head of the European Defence Agency, Federica Mogherini said: “Cyberspace threats do not know of national borders. Cooperation among Member States but also at European level is therefore essential. Europe is stronger when it tackles threats together, in a common and coordinated approach. And this is exactly where this Memorandum of Understanding is key and where the added value of the European Union lies: working together, joining forces, putting the experiences and the knowledge of all at the service of our citizens' security. ”
Vice-President for Digital Single Market Andrus Ansip said: “We can face cyber threats successfully if we have in place a functioning exchange of information, we have strong technical capabilities and we work on basic cyber hygiene. Better cooperation between these EU agencies will lead to this result. ”
Commissioner for Migration, Home Affairs and Citizenship Dimitris Avramopoulos said: “The threats against both our physical and virtual worlds are becoming increasingly connected. This is why increasing cyber security is one of the priorities of the European Union. But we can only do this effectively through stronger cooperation and joint actions, where our operational agencies, like Europol, can play a critical role with the expertise they bring to the table in support of our Member States. ”
Commissioner for the Security Union Julian King said: “The cross-border nature of the cyber threat means that cooperation has never been more important. This improved collaboration between ENISA, EDA, EC3 and CERT-EU will help us to strengthen our cyber resilience, build effective deterrence and help deliver credible cyber defence and international cooperation. ”
Commissioner for Digital Economy and Society, Mariya Gabriel said: “Trust and security are key components of the digital economy and society. The EU agencies should lead by example. Only by working closely together will we have a chance to mitigate the cybersecurity risks.”
Prof. Dr. Udo Helmbrecht, Executive Director of ENISA said: “ENISA welcomes the opportunity to work closely with our partner organisations. Cybersecurity is a shared responsibility, and it is only by cooperating closely with all relevant stakeholders that the EU has a chance to address cybersecurity challenges.”
Jorge Domecq, Chief Executive of the EDA: “EDA supports Member States in the development of their defence capabilities. As such, we also act as the military interface to EU policies. Today’s Memorandum of Understanding is an important step towards increased civil-military cooperation and synergies in the area of cyber security and cyber defence.”
Steven Wilson, Head of Europol’s European Cybercrime Centre (EC3): “This MoU illustrates how a safe and open cyberspace can only be achieved through enhanced cooperation and commitment. Through their participation, all parties involved demonstrate that they are willing to join forces and recognise that together we can provide the necessary response to cyber related threats. From EC3, we welcome the opportunity to enter a new era of working together with our MoU partners and are delighted to share our expertise and experience.”
Ken Ducatel, Acting Head of CERT-EU, said: “The EU institutions, bodies and agencies rely on the specialised skills and tools in threat intelligence and incident response of CERT-EU. But, we don’t maintain these capacities by acting alone. That is why acting together with our peers and partners in the other signatories to this Memorandum is so important.”
The 2014 Cyber Defence Policy Framework called for the promotion of civil-military cooperation and synergies with wider EU cyber policies, relevant EU institutions and agencies as well as with the private sector. ENISA, EDA, EUROPOL and CERT-EU began initial discussions in 2016 which eventually led to this milestone signature. The principles behind this Memorandum of Understanding are fully in line with the implementation of the Joint Communication on Cyber issued by the High Representative and the European Commission in September 2017.
More information:
The Taleban look to be preparing for a new onslaught on Andar district centre. The name ‘Andar’ is still full of political resonance, gained in the summer of 2012 when the Taleban were suddenly and swiftly pushed out of a large part of the district. That counter-insurgency in an insurgent stronghold was styled the ‘Andar Uprising’ and was promoted enthusiastically by the government and United States military they hoped it marked the start of a wave of popular revolts against the Taleban. But by late last year, the last areas captured by the uprisers in 2012 were lost back to the Taleban. The government now controls just a tiny sliver of land and that precariously. Roads to the district centre are cut off and residents are preparing themselves for a new onslaught. In this latest in a series of dispatches on the Andar uprising published by AAN since 2012, Fazal Muzhary and Kate Clark consider why the government has failed so badly in Andar and what it tells us about the attractions and perils of raising ‘community defence forces’.
This piece draws on earlier AAN research on the Andar uprising (1) and subsequent developments, as well as a range of new interviews with locals (both combatants and civilians) and international officials. Revisiting Andar is an opportunity to put events there in context, in the light of a research project* which in part is looking at community defence forceswhat makes them successful or not. Andar is a good example of a community defence force which failed. Although the uprising group and the Afghan Local Police (ALP) which it soon mainly turned into were initially successful in their fight against the Taleban – with strong Afghan National Security Forces (ANSF) and international military backing – they ultimately failed to hold territory. In addition, introducing a local counter-insurgent element led to extreme violence, producing the opposite outcome from the ‘population protection’ mantra that has been used to justify the mobilisation of community defense forces.
This dispatch can be read alongside a forthcoming analysis of the ALP in Yahyakhel district, in neighbouring Paktika, looking at why it has been highly successful.
The arc of government failure in Andar
All the signs are that the Taleban are preparing to launch a fresh attack on Andar district centre. On 4 May, as witnessed by one of the authors, they used an excavator to dig a hole on the outskirts of Ghazni city on the main, asphalted road which leads to Andar and then on to Paktika. The highway was already blocked in the other direction, meaning Andar was already cut off from the Afghan National Army (ANA) base at Chahardiwal,to the east of the district headquarters. With a third road also blocked by the Taleban, supplies can only now reach the district centre by air, with government helicopters vulnerable to insurgent rockets. A government operation to open the Ghazni-Paktika highway began on 14 May, but failed to get beyond the outskirts of Ghazni city. Residents in Andar told AAN the insurgents have distributed letters to officials offering amnesties: if they surrender their weapons and go home, they have been told, they will not be harmed. Provincial council member Ahmad Faqiri warned that “If the district doesn’t get supplies and reinforcements, there will be a disaster.”
Local people have been waiting, as they have all winter, for the Taleban to again attack the district centre. In October 2017, it suffered an intense offensive and three-day siege – after the Taleban had blocked supply routes. (2) 60 Afghan National Police (ANP) and ANA managed to hold out against a Taleban force of 300 for three days, from 17 to 20 October. (Significantly, there were no longer any indigenous uprising or ALP forces still operating, only a small number of Shinwari ALP from Nangrahar, who did not fall back from their posts just outside the district centre on the road to the ANA base, to defend the town centre.) The Taleban onslaught was broken only by intense air strikes by US forces and the eventual arrival of ANSF reinforcements, including commandos. More than 70 people, including civilians, security forces and Taleban, were killed and injured in the attack. (3) The Taleban also captured pretty well the last of the territory they had lost to the uprisers in 2012. Government forces were left controlling just the district centre – now badly damaged – and four nearby villages.
Since the three-day siege, the ANA was making occasional forays against the Taleban and the insurgents were lobbing occasional rockets at the district centre and attacking ANP checkposts in the nearby villages. However, the government failed to regain control of any of the area round the district centre. Few people from outside town were risking coming to thebazaar and only three of the 16 families who fled the district centre and surrounding villages in October returned. The Taleban warned residents they would attack Andar again. They now fear that is imminent.
What happens in Andar is strategically important. It lies on the east-west road joining Ghazni city (37 kilometres away) to Paktika province and forms one of the ‘gateways’ to Ghazni’s provincial capital. The main Kabul-Kandahar highway also passes through the western part of the district; insurgent control makes this vital road vulnerable to closure. The district also hosts one of the most important madrassas in Afghanistan: Nur ul-Mudares has, for decades, served as the religious hub for the whole south and south-east of Afghanistan, supplying top-level mullahs, madrassa teachers and imams. Andar also retains its political significance because of the 2012 uprising. Six years ago, it was the bellwether for those who hoped popular uprisings would lead to a routing of the Taleban. This dispatch looks at the reasons for the government’s failure there.
The nature of the 2012 uprising
Andar (also known as Shelgar) district went over to the Taleban insurgency very early on after the collapse of the Taleban regime, with young men who had mainly been madrassa students when the Taleban were in power taking up arms against the new government and its foreign backers as early as 2003. By 2012, Andar had been solidly held by the Taleban for years. Yet in May and June of that year, a new, local group of counter-insurgents formed. They called themselves De Melli Patsun Ghorzang (the National Uprising Movement), a term soon used for similar groups elsewhere in the country; members called themselves patsunian. In a rapid and unexpected campaign, the uprisers gained outright control of 46 out of the district’s 480 villages and stopped or reduced Taleban influence in others, so that the insurgents’ freedom of movement was hampered and constrained in about half of Andar (see detail here). The most immediate effect of the change of control was that schools, which the Taleban had closed in response to a government ban on unregistered motorcycles, which were being used to launch attacks, were re-opened.
Many in the media, in government and among Afghanistan’s foreign backers hailed the event as a ‘popular uprising’. “Villagers take the counterinsurgency into their own hands,” reported The Economist. Radio Liberty described how “[a] group of angry Afghan villagers have got the Taliban scrambling after they mounted an unlikely rebellion against the insurgents in eastern [sic] Afghanistan – and won.” Meanwhile the Washington Times wrote that, “Fed up with the Taliban closing their schools and committing other acts of oppression men in a village about 100 miles south of Kabul took up arms late last spring and chased out the insurgents with no help from the Afghan government or U.S. military.” Influential American commentators, Frederick and Kimberly Kagan writing in the Wall Street Journal, differed in the detail, but not in their enthusiasm:
The Taliban attempted to crush this nascent resistance. But local fighters supported by NATO and Afghan forces defeated them, sending shock waves through the Taliban leadership and the Afghan government… As a result, many villages across Afghanistan are now modeling the “Andar Uprising,” by which they mean forming anti-Taliban groups that seek the help of NATO and the Afghan military. This phenomenon is not as widespread or pivotal as Iraq’s “Anbar Awakening” in 2006-07, when Sunni tribesmen helped turn the tide against al Qaeda-backed insurgents. But it is extremely important as a harbinger.
Yes, as AAN research at the time found (read it here), the ‘uprising’ in 2012 was far messier than generally reported. This was an intra-militant struggle, rather than a case of popular resistance. The revolt was initiated by a group of young Hezb-e Islami members who had joined the Taleban. Their decision to move against their comrades was certainly buoyed up by widespread local discontent with the Andar Taleban’s particularly harsh rules; these included the widespread closure of schools and bans on development work and visiting the district centre, including the Mirai bazaar, which is located there. They had also forbidden mullahs from giving Islamic funerals to those killed by the Taleban because they worked with the government or were ‘spies’. Calling the rebellion a ‘popular uprising’, though, in the sense of an uprising organised and carried out by local communities was a misnomer. Local people, aside from the Hezbi fighters, were not actively involved.
Moreover, a variety of patrons, both pro-government politicians and US special operations forces, rushed to support the counter-insurgents, leading to splits, including from the original rebels who said they were with neither the Taleban nor the government. They condemned the politicians trying to harness the uprising as Mafiosi. The various splinter groups which emerged had different loyalties and chains of command, recruiting young men largely through informal networks. One international working in Ghazni at the time described the rivalries among those seeking to control and benefit from the force, including between Governor Musa Khan Akbarzada and Ghazni-born former governor and then NDS Director Asadullah Khaled. (4) “They spoiled the dish from the beginning,” he told AAN. “There was no chance of a genuine, endemic, local rebellion from the bottom up that had legitimacy from local people to develop.” In August 2012, AAN reported that among locals, “feelings are mixed about what has happened.” Some celebrated the re-opening of schools and of the Mirai bazaar and the distribution of aid. Others were already worried:
…people had more optimism for the change in the beginning when it was mainly ordinary local youths who were fighting. Later, when they saw former commanders getting involved and when members of the arbakai started harassing some people who were or had been sympathetic to the Taleban, concerns started to overcome the optimism. Many now worry about the way power changed hands and fear a new phase of factional violence could be looming: internecine conflict among the Andar tribespeople would have long-lasting repercussions.
In that same August 2012 dispatch, we reported that the government had refrained from officially adopting the Andar ‘rebels’ as an ALP unit under the Ministry of Interior, “[T]his would seem prudent,” we said, “given the lack of tribal cohesion among the Andar and the local residents’ hesitance [sic] to support a government-designed plan.” However, by October, a new ALP unit had been established.
Why set up an ALP in Andar
According to both ALP regulations and the ethos of the programme, local involvement is an integral and essential element in setting up ALP units. However, according to local people, they were not consulted on whether they wanted an ALP and if they had wanted it, who they wanted to serve in it. (For detail on the regulations and how they were all too often ignored, see here and here, especially pages 12-17). One respected local businessman, who asked not to be named for security reasons, told AAN that most locals were simply unaware that the ALP was being set up in their district. He said they only learned about it when they saw armed men wearing ALP uniforms in the district centre. “The only difference between ALP and the upriserswas the uniform,” he said, “The upriserswore no uniform.”
Rather than community demand driving the decision to establish an Andar ALP, various Ghazni politicians were pushing for it and, most significantly, the US military wanted it – or at least, the leadership did. There was actually a great deal more scepticism about the wisdom of supporting both the uprising and the ALP among internationals working in Ghazni, as several have described to AAN (all asked not to be named). Along with another international working at mid-level in ISAF HQ, they described a disconnect between what the international military leadership wanted to believe was happening and what those on the ground knew was happening.
“News of the uprising had reached ISAF HQ,” remembered the international working at ISAF headquarters, “They were desperate to find anything that looked like progress or demonstrated a change in the engagement in the population to support their mission.” In Ghazni itself, though, he stressed, both civilian officials and military officers did have an accurate idea of what was going on.
We had fidelity of information at the tactical and operational level. There were smart people there. They had been able to work out what was actually going on. The decision-making at the centre was more influenced by politics in the home nations… and there was a real desire to represent [the uprising] as progress. Headquarters wanted a good news story. They had a cartoon view of the uprising in their heads: the Taleban were trying to close the schools and the kids had risen up and said, “We want our education,” because they wanted jobs and a future… I would characterise some of the thinking as cognitive dissonance.
The leadership, he said, did not want to hear what was actually going on in the field. He cited an example of a senior officer refusing to relay a report from the field upwards because it contradicted what the command wanted to hear.
Another interlocutor based in Ghazni described US civilian officials in Kabul as “gushy and breathless about how [the uprising] was going to change everything.” He said they thought “the locals are fighting the Taleban. There will be a genuine, local endemic rebellion and we’ll be able to leave. They were blind. They didn’t see what was happening because they didn’t want to see what was happening.”
At the time, the man in charge of US and NATO forces was General John Allen. He had been a key commander overseeing the Anbar Uprising in Iraq. The similarity of the place names – Andar and Anbar – and the hope that the Sunni Awakening could be replicated as a counter-Taleban counter-insurgency in Afghanistan appears to have been too tempting. Despite advice to the contrary, which interviewees told AAN he was given, Allen publically drew parallels between the two utterly different uprisings. “This is a really important moment for this campaign,” he told Foreign Policy magazine (quoted here: “because the brutality of the Taliban and the desire for local communities to have security has become so, so prominent—as it was in Anbar—that they’re willing to take the situation into their own hands.”
Those in the US military tasked with setting up the ALP found willing partners in the various politicians from Ghazni who had already been quick to try to leverage the uprising and any new force, especially if it was internationally-backed, to enhance their own power and gain resources (armed men or money). These includedGovernor Musa Khan and NDS Director Khaled, and two jihadi-era rivals, both of whom were former MPs and commanders, Khial Muhammad Hussaini (Harakat-e Enqelab-e Islami) and Abdul Jabar Shelgari (Hezb-e Islami). More details about the lobbying and rivalry of commanders, politicians and notables and how this created splits in local forces along sub-tribal lines and between mujahedin factions can be found in AAN reporting from August and September 2012 and April 2013.
Recruitment into the ALP
Most, but not all uprisers became ALP. There was also fresh recruitment. If the ALP had been a genuine community force, one would have seen fathers introducing their sons to the force, or a part, at least, of the community coming together to organise some of their young men to join. Instead, the view from the ground, as described to AAN, was that most of the recruitment was based on friendship relations between peers. The local businessman referred to earlier described how the ALP would “immediately recruit” those put forward by existing members “and give them weapons” with little scrutiny as to their competence or checking in with the community. He said many of those recruited were jobless youths recently expelled from Pakistan. They joined the ALP not out of a desire to liberate their communities from the Taleban, but, said the businessman, because they needed jobs, and because being in the ALP and having a weapon would make them powerful men in their villages. A resident of one of the villages where the uprising started, who also asked not to be named, said “The actual tribesmen were not consulted about the creation of ALP. Neither were they told to introduce men to serve in it.” Rather, he said, in those villages already controlled by uprisers, youth would individually decide to join the new force. “Every day, we would learn that this or that person had joined.” The UNAMA Protection of Civilians annual report for 2012 also said that “community members reported dissatisfaction with… the ALP recruitment processes.”
Another perspective on ALP recruitment in Andar comes from military academic, Matt Dearing, who was a member of the civilian Human Terrain Team in Ghazni advising the US military at the time and is now an assistant professor at the National Defence University in Washington DC. (5) He described a local figure called Khalil Hotak, a former Jamiat-e Islami commander, influential political activist and head of a council of tribal elders from across Ghazni, the Community Salvation Council, as having carried out training and recruitment of “resistance fighters into the ALP.” The US special forces were then responsible for vetting and training those put forward.
KhalilHotak enjoyed good relations with the US military from 2001 onwards and, said Dearing, was often at the US military base in Ghazni. (He has also been accused of grabbing thousands of acres of state land in Ghazni province (see for example, this report from Khaama news agency.) In conversation with AAN, Dearing said the men Hotak put forward were mainly, but not all local – they also included people from Kabul and Wardak. Nevertheless, they were men he could vouch for. He was, said Dearing, “central to ALP recruitment.”
This outsourcing of some ALP recruitment to Hotak appears to have represented the extent of ‘community involvement’ in the creation of the new force. As to why the US military chose to outsource this crucial task, Dearing said, “Khalil Hotak was someone who was able to make himself appear to have a lot of influence throughout Andar district. His name was always there or brought up in key leader engagements, he would come to the base and bring food and host dinners for the special forces. He was always lobbying, for his own interests possibly, or the interests of the greater community, I don’t know for sure.”
Normal procedures for recruitment may been shortcut because, as Dearing described it, US conventionalforces – three companies – were preparing to draw down in the province (6) and special forces teams had “limited intelligence on demographics and the needs of the rural population, let alone the Uprising.” At the same time, however, the US military in Ghazni were under immense pressure to mobilise ALP rapidly:
At an individual level, there were US commanders who realised and wanted to know if there were abuses going on, realising that there needs to be accountability and disciplinary measures. There were people at battalion or company level in Ghazni, dealing with these things on a daily basis and meeting uprising leaders, but they were getting a lot of pressure from above to make [the ALP] happen. [They were told]: “There needs to be ALP. We have to have X number of ALP by such and such a date.” The need to get the numbers up took precedence over micro-level local concerns.
Whether a genuine community defence force – defined as one where the community is consulted, agrees to the force and has some control over membership, and recourse if that force behaves badly – could have been established in Andar is, of course, unknown. The district is almost completely mono-ethnic and mono-tribal – practically all the inhabitants are Pashtun Andars. However, there are many sub-tribes and Andars are notoriously disunited – local people use the phrase “as disunited as Andar” proverbially. Moreover, tribal solidarity had worsened both during the reign of the brutal mujahedin commander Qari Baba in the 1990s – himself an Andar and fighting with Harakat – and the Taleban insurgency.
At the same time, however, Andar people have proved their ability to organise and take collective action: in 2013, for example, gatherings were held on both sides of the frontline to consult on and set bride prices because they had become too high (see here). In December 2013, local people also gathered to discuss building bridges (see here) across the river passing through the district. These gatherings were organised by local people without government involvement. This level of community coordination and collective action suggests that consultation on the ALP might have been possible. (Whether the community would have then supported such a proposal is another question – by this time, the indiscipline and non-accountability of the armed young uprisers may well have made the idea of setting up a formal local defence force unattractive to many.)
Whether possible or not, it appears neither the state nor the US military sought to consult the population. Instead the first introduction of the idea to the community wasa ceremony held on 21 October 2012 to formally announce the formation of the ALP in Andar district, and introduce the members. Even then, few Andar elders were present to witness Ministry of Interior officials, including the head of the ministry’s ALP section, inaugurating the first unit. All those who came were from areas under the complete control of the uprisers,and almost all were already or would soon became involved in the ALP, either directly or through their sons. (7)
Why two forces?
Andar got an ALP, but it also retained uprising forces. In April 2013, they amounted to a few dozen fighters, mainly were from the original Hezbi group of rebels (see reporting here and here which also describes a third group who left altogether out of disgust at their movement being ‘hijacked’ by the government). Some of those who stayed as uprisers gave personal reasons for not joining the ALP – they said they still had a separate and different cause – but there were also ‘demand’ factors causing the continuation of two separate forces, argues Dearing. “While patrons offered strong incentives to join ALP, they also continued to incentivise… [the uprisers] by maintaining alliances, weapons and resource provisions, and conducting joint operations.” For them, “… the necessity for a local, informal security response led to continued state support to predatory paramilitaries in Ghazni.”
Both the ALP and uprisers in Andar were touted by government and the US military as ‘community defence forces’, but actually they answered to other interests and that, in the end, had an impact both on their behaviour and on the weakness of the community in getting them held to account. Locals rarely differentiate between the two forces, referring to both (as do the Taleban) by the now pejorative term, ‘arbaki’ (here meaning an undisciplined and unruly pro-government militia)
Andar versus Andar, an intensification of violence
In autumn 2012, we reported that Andar had become “one of the most heavily militarised zones in Ghazni.” As well as the newly formed ALP, there were the remnant uprisers – different armed groups, each with its own chain of command, clustered in the villages immediately to the west and south of Mirai, where the uprising had started. The ANP and ANA deployed forces – UNAMA described a “strong” ANSF tashkil– to the district centre, as did the NDS (it has paramilitary units). US troops, including special forces, were also operating in the district. The ALP, formally under Ministry of Interior command, had strong and influential local patrons. As to who the remnant uprising forces answered to, that was not clear, but they were supported by and carried out joint operations with NDS and US special forces. Taleban were also still present, “roaming freely and calling the shots in most parts of the district,” we reported that autumn. Locals told UNAMA there had been “an increase in Taliban forces from outside Afghanistan in the area following the uprising.”
The mobilisation of anti-Taleban local forces led to an intensification of violence. One estimate of the total number of Andars killed from all sides in 2012 was 102, which would have made it the bloodiest year for the district since the start of the insurgency. (8) The impact specifically on civilians was documented by UNAMA in its 2012 annual Protection of Civilians report: 45 civilians were killed or injured in Andar that year, the majority of whom were directly or indirectly related to the uprising. “While the uprising movement did not involve the direct targeting of civilians,” UNAMA said, “the presence of a new fighting force, an increased presence of ANSF counter-insurgency activities and the establishment of ALP combined with increased numbers of Taliban sent to counter the uprising, all contributed to civilian casualties.” (9) The high number, it said “highlighted the frequency of violent clashes” in Andar.
Over the winter of 2012/13, the militias expanded their foothold by capturing more villages from the Taleban in the south of the district and to a small extent in the east. These villages had already been abandoned by the Taleban and had been serving mostly as buffer zones. However, the majority of villages in Andar never came under ALP or uprising control.
In spring 2013, both government and Taleban were promising to eliminate the other’s forces in Andar. By November of that year, AAN was reporting that the violence there had become “increasingly savage,” with rough estimations by local elders and notables of more than 300 people killed since the start of the uprising, far exceeding all the dead of the conflict between summer 2003 and summer 2012. “More shockingly,” AAN wrote, “the conflict has spread not only in numbers, but in the quality of the violence, with a widening of targets and tactics.”
As a future dispatch looking at Taleban attitudes towards community defence forces, which uses Andar as one of its case studies, will show, Taleban violence against the ALP and uprisers was extreme during this period. No quarter was given (by contrast, the Taleban allowed ANA and ANP to withdraw or surrender). The Taleban also made attempts to infiltrate the ALP and get ALP members to defect in order to carry out mass ‘green-on-green’ killings of local policemen. They continued to carry out reprisals against civilians associated with the ALP and uprising forces. UNAMA had already reported on this in 2012 when it documented five targeted killings of pro-uprising civilians. In November 2013, a roadside bomb targeted a van of wedding guests travelling in an ALP area. It killed 19 people, almost all women. Locals lynched a young man assumed to be with the Taleban whom they caught allegedly running away from the scene and beat him to death. As we wrote “…the belief that the ‘other side’ would want to kill female wedding guests (and it is hard to think of a killing with a stronger taboo), stems from the mounting aggressiveness and hatred perpetuated by both sides involved in this conflict, an enmity which has been partly fuelled by rival mullahs.”
2013 also saw a ban by both sides on the Islamic burial of enemy combatants. Six clerics were killed that year for violating this ban, two on the Taleban side, two on the ALP side and two others, whose ‘affiliation’, if any, was not clear. In November 2013, a Taleban mullah was reported to have issued a ruling that everyone in ALP villages was a target (he was subsequently killed by the ALP). A year previously, locals had reported to UNAMA that a Taleban mullah had issued a fatwaagainst members of the uprising, for them to be killed and their wives taken. The judge was killed in November 2012 in a search and raid operation by ANSF and international military forces – the Taleban condemned the killing of the “local imam” by arbakiand international forces.
Both sides fought hard. The Andar ALP was one of the most robust ALP stood up, tough in battle, and also enjoying a lot of support, at least initially, from both US and Afghan forces. The remnant uprising force also received money and support, especially from the NDS and US special forces. Both fought in joint operations with other Afghan and US forces. Like the Taleban, ALP and uprisers carried out reprisals against those they believed belonged to or were sympathetic to the other side; as detailed below, the victims included civilians.
Abuses by the ALP and Uprisers
Only in the very early days of the uprising were fighters relatively well disciplined. As we reported, at first, the extent of uprising fighters bothering locals was to ask them to provide food or money for their basic survival. By October 2013, that had already changed, with residents reporting armed youth arresting people coming from Taleban-controlled villages and those whom they suspected of being pro-Taleban. Frequently, the detainees were released only after paying money or being robbed of their goods. Some were beaten. Such abuses, particularly of the population of newly conquered villages, continued into 2013.
Over the next two years, reports of serious abuses by both the ALP and remnant uprising groups persisted. In 2014, AAN detailed ALP units extorting money from detainees. For example, the commander for Andar district centre, Rohullah, arrested a mullah, Faiz Muhammad from Tut village, in the district bazaar in late May 2014, beat him and freed him only after he paid 180,000 Pakistani rupees (roughly 1,800 USD). In early June 2014, ALP members detained another mullah, the imam of a local mosque in Mirai under the pretext of interrogating him again about his alleged relations with the Taleban. “Such detentions and extorted payments for release,” we wrote, “resulted in a widespread frustration among locals, especially after customers, concerned about these developments, stopped coming to the bazaar.” In November 2015, Killid reported that ALP men arrested and killed a well-digger, Shah Wali, from the Taleban-controlled village of Mehman, near an ALP post in the Kajera area and then ran their Ranger vehicle over his body.
As for uprising forces, UNAMA devoted a whole section of its 2014 protection of civilians report to those in Ghazni. (It analysed ALP nationally and in a separate section.) It reported “an incident of collective punishment and alleged crimes involving more than 40 civilians that involved severe beatings, including with metal chains” which had been carried out in January 2014 in Andar. Al-Jazeera detailed the alleged killing of three people in Andar in June 2014 by an uprising commander, Abdullah. The allegations were given weight by then head of the United Nations human rights unit in Afghanistan, Georgette Gagnon, saying they had “investigated and verified allegations of extrajudicial killings of three men by a pro-government militia.” According to al-Jazeera, US Special Operations forces and Afghan National Army commandos had carried out a night raid on Alizai village on 1 June and detained about a hundred men from the village in a compound. Late on the following morning, it said, a mix of uprisers and ALP commanded by Abdullah came by. They returned that afternoon and took away three of the detainees who were later shot. Al-Jazeera quoted Abdullah saying, “‘[I]t was a raid and I caught them… If anyone is saying these were civilians, that person is pro-Taliban, a Talib himself, or is spreading Pakistani propaganda.’”The following year, The New York Times reported on a father accusing one of Abdullah’s sub-commanders of having killed his son, a 13 or 14 years old, in January 2015 after questioning him about roadside bombs. Abdullah was quoted by the paper calling the US special forces “my brothers.”
Locals did complain. Abusive behaviour towards the residents of newly captured villages in early 2013 prompted a rare demonstration by hundreds of residents of Andar and the adjacent Deh Yak district (where the first squad of 50 ALP had been deployed) on 16 March against “the arbaki”. “The people of these districts,” we wrote had “probably never publicly protested against local actors before. The lack of a culture of demonstrating and local disunity makes it difficult for them to even think of coming together for such an untraditional action and yet they did.” In 2014, residents of Gelan district managed to get one abusive uprising commander removed, reported UNAMA, “following interventions with Afghan national security forces and Government authorities.” It observed, however, that “the removal did not improve civilian protection.” In 2016, Andar women went to Ghazni city to protest (see a video here against the abusive behaviour of arbaki who, they said, were forcefully breaking into homes and abusing women. They said they were fed up with this behaviour. However, little, if anything, was ever done.
Command and control of ALP and uprisers
Both the ALP and uprisers were abusive towards at least some members of the local population. Trying to judge if one group was worse than the other is difficult. People in Andar do not differentiate between the different types of ‘arbaki’, and there is also not enough publically available data on abuses from sources like UNAMA to properly assess. On paper, the ALP should have had better command and control, given that the force had a legal underpinning, clear official lines of command from district and provincial police chiefs and a mechanism in the Ministry of Interior for dealing with ALP infractions. According to the ethos of the programme, ALP units should also have come under community control.
By contrast, there is no basis in Afghan law for the existence of uprising groups and it has never even been clear who they answer to in practice. The Andar uprisers’ patrons, primarily the NDS and US special forces (both themselves less transparent and less accountable than the ANP or the regular US military), supplied them with weapons and funding. However, UNAMA said, in 2014, the uprisers in Ghazni province had perpetrated crimes against the civilian population with impunity. The government, it said, had undertaken “no investigations or remedial efforts.” In the same report, assessing measures undertaken by the Ministry of Interior to investigate and prosecute wrongdoing by the ALP nationally, it said these had improved marginally and from a very low level (nationally, the crimes the MoI had investigated ranged from murder and arbitrary detention to illegal search operations, forced evictions, extortion and mistreatment. (10)
Dearing thinks the ALP was less abusive than the uprising forces and puts this down to better control from both the community and Ministry of Interior, although he provides little evidence for this in his writing. (11) In conversation, he argued that there was a freedom that came from being a member of the uprising force:
They could be a little more brutal and predatory, more than the ALP was allowed to be. Uprisers could do what they wanted and still get resources and funding from the NDS. American and Afghan special forces and NDS could always go to uprisers for clearing and offensive operations… The ability of those forces to not hold them account, to look the other way was certainly there. That’s why you use these types of militias and paramilitaries, partly because they are cheaper and partly because they are not attached to you, but they do your bidding to a certain extent. It’s a dangerous game. We could always turn our backs and say, ‘We don’t control them.’
Dearing takes this subject onto a new and important theme – how tight discipline and protection of the civilian population may actually notbe in the core interests of a patron, or at least not interesting enough for the patron to try to deal with. The uprisers, Dearing writes, were able to bring their “unique intelligence capabilities” to bear primarily because they “resided outside the chain of command.” Indeed, the support given by patrons to the uprisers to act as a covert counterinsurgency force gave them “greater latitude to operate with impunity.”
After 2014 – The Taleban gain territory
From 2014 onwards, it was the Taleban who saw military gains in Andar. The onslaught from them on the Andar ALP between 2012 and 2014 had succeeded in breaking the momentum of the counter-insurgency and the government began to lose territory, sometimes one or two villages at a time, sometimes many more. In February 2016, for example, they lost the villages of Khar, Karpal, Sanginaka, Mahkam, Rustam and Manar all at once. This change in fortune also coincided with the ANSF taking responsibility for security as international forces withdrew.
ALP and uprisers themselves blame government negligence for their decline. For example, in this opinion piece written by a leading member of the uprising force under a pseudonym on 5 August 2013, the author says lack of state support was undermining their ability to fight, to recruit and attract the population to their cause. Four years later, ALP demonstrating in Ghazni city in January 2017 told Pajhwok that, “The Afghan government promised us money and weapons, but provided us nothing.” There was no significant ANP and ANA deployment to the villages under threat. In the face of Taleban violence, militiamen had no sense that the government was protecting their backs.
As a forthcoming dispatch on Taleban attitudes to the ALP will detail, the movement’s tactics also changed around 2014. Although the Taleban military campaign against the militias in Andar continued, they were no longer regarded as their most important enemy. The insurgents started to use a more ‘softly softly’ approach in Andar and elsewhere, trying to persuade ALP men to switch sides, offering amnesties and trying to address the grievances of communities which may have supported the ALP.
Significantly, the Taleban have not undertaken acts of reprisal against those living in re-captured villages.That pragmatic approach has extended even to former ALP members and their families who sought amnesties and chose to stay in their areas. For example, according to businessmen speaking to AAN in 2017, when Taleban fighters detained two former ALP men, they let them go after a few days. Even though active ALP and uprisers remained targets, we also saw in 2016 the Taleban handing over the body of an ALP man whom they had killed after forcing him from his car on the Kabul-Kandahar highway. Such behaviour would have been unimaginable at the height of ALP-Taleban hostilities. Andar went to the brink of an abyss in terms of the ferocity of intra-tribal violence. It seems that for now, at least, the Taleban have realised that better behaviour towards local people, including those who went over to the other side, can be militarily useful to their campaign in the short-run – for winning back territory – and thelong-run – keeping that territory.
As the ALP and uprisers in Andar were gradually pushed back, their numbers also dwindled. In early 2017, then ALP commander Baz Muhammad said there were still 50 or so ALP and uprisers. About a dozen defected or surrendered to the Taleban in the autumn of 2017. Others went to Ghazni city or Paktia or Khost provinces where they started businesses. By October 2017, none of the original group of uprisers or ALP were left. The only serving ALP in the district were Shinwaris, drafted in from Nangrahar province (an exception to ALP regulations); this was reported by locals and confirmed to AAN by an official from the Ministry of Interior, who asked not to be named, and said it had been necessary because “locals aren’t interested in serving.”(12) These ALP men are holding a couple of security posts on the road leading to the Chahardiwal area which houses the ANA base.
Baz Muhammad, who ended up commanding most of the remaining uprisers,himself finally left the ranksout of frustration with the formidability of the task of ensuring security in a difficult area, and because of the lack of popular support. The ALP and uprisers, he told AAN, found themselves increasingly surrounded by the enemy and trying to operate in a terrain in which the population – after experiencing different types of armed actors – had lost interest in actively supporting any brand of people under arms. Baz Muhammad joined the ANP in February 2018 and was assassinated at his post earlier this month. (13)
Assessing the Andar uprising and the ALP
The uprising in Andar district failed to hold the territory it captured. In terms of the human cost, according to provincial council member and brother of an early ALP commander, Amanullah Kamranai, more than 700 uprisersand ALP men have been killed since 2012. An unknown number of civilians, ANSF and Taleban also died in the violence, which was particularly intense between 2012 and 2014.
Several themes emerge from the arc of failure in Andar. First, changes in the battlefield in Afghanistan can be rapid and unexpected. Winning over local civilians appears to be one crucial means of making gains long-lasting. In insurgencies, the support of civilians can tip the balance between warring parties, for example, if they chose to give – or withhold – tip-offs and intelligence. By the end of summer 2012, abusive behaviour by uprisers, the mercenary behaviour of local politicians, and probably, as well, the local forces’ alliance with foreign forces, was already limiting their potential support from local communities. Whether a popularly-based ALP could have been established then is unknown: the force that was stood up was created without consulting or involving local communities. Nevertheless, the ALP and uprisers could still manage to draw on the support of part of the community because of the importance of extended family networks and this made them a threat to the Taleban – as we wrote in October 2013:
If the Taleban had been an external force without indigenous roots, they would have easily been swept out of the district by such a powerful local militia. However, the Taleban have established support among a considerable segment of the society and it is this entire segment of society which has found itself the enemy of the ALP. At the same time, the ALP also has local support and those within the community who are or are perceived to be ALP-aligned now find themselves the target of the Taleban. The result has been relentless bloodshed perpetrated by both sides and a polarisation within the Andar tribe.
That last sentence is also important. A second theme emerging from events in Andar is that introducing a local element into a counter-insurgency is risky, in terms of the prospects both for stabilisation and protecting the population. Although the ALP and uprisers fought hard against the Taleban, they never managed to get enough local support to become the dominant force in Andar, which meant their presenceworsened the violence and reduced protection for civilians. The consequence was an intensification of the conflict, and, ultimately, no clear counter-insurgency win. Indeed, the situation eventually reverted, after a lot of blood was shed, to the status quo ante.
Inserting a local counter-insurgent force also localised the conflict to a much greater degree, producing a more brutal and intimate type of violence, something likely to be more pernicious and de-stabilising in the long-term. Before the 2012 uprising, the war in Andar had mainly been fought out between Taleban with a local recruitment base and foreign soldiers, with some Afghan ANSF. After the summer of 2012, it was Andar versus Andar. Those fighting knew each other and each others’ families and the conflict was fought out in their own villages and over their own lands.At the height of the struggle between the Taleban and the militias, the level and nature of the violence was worse than anything seen before, even under the ruthless rule of Qari Baba. The arming of two sections of the community – Taleban and uprisers/ALP – poisoned intra-tribal relations among the Andar and led to extreme acts of reprisal. That Taleban have not carried out revenge attacks in re-conquered villages in Andar – presumably for pragmatic reasons – can only be seen as fortunate.
The local nature of forces such as the Andar ALP and uprising group is what makes them attractive to various patrons, both Afghan and international. They bring tactical advantages – local knowledge of the terrain and the community, and clan and family networks. In the case of Andar, in the early years of the uprising, these paid dividends in terms of territory taken. Raising local forces may therefore be good for those planning the counter-insurgency, while having bitter consequences for the community. In Andar, regardless of whether the ALP or uprisers had managed to retain the territory they won, the long-terms harm of raising local militias was clear from very early on.
Finally, although the ALP and uprisers were ‘advertised’ as community defence forces, the communities did not control them. There was very little evidence, either, of any proper accountability of uprisers (by the NDS) or ALP (by the Ministry of Interior), or by US special forces of either. The indiscipline of these militia forces cost civilians dear, in terms of greater insecurity and criminality. Such indiscipline may have seemed a reasonable price to pay by patrons wanting a robust and dependable counter-insurgency ally against the Taleban, which they could hold at ‘arm’s length’. However, that predatory behaviour ate away at civilian support, ultimately weakening the force. Even though the ALP and uprisers could count on some backing from some of the community, it was never going to be enough to enable them to hold territory, however well they fought, if they also abused the citizenry.
That the Andar experience of ‘community defence forces’ is not inevitable can be seen by looking at what happened in Yahyakhel district in neighbouring Paktika provinces. The establishment of the ALP there sheds light on how it is possible to stand up a good local force, which both protects civilians and reduces violence, a force that is seen as legitimate by the ‘host’ community. This will be explored in a forthcoming second dispatch.
Edited by Erica Gaston
* This research for this dispatch is part of a joint, three-year project by AAN, the Global Public Policy institute (GPPi) and the American University of Iraq, Sulaimani. The project explores the role and impact of militias, local or regional defence forces and other quasi-state forces in Afghanistan, Iraq, and Syria, including mechanisms for foreign assistance to such actors. Funding is provided by the Netherlands Research Organisation.
(1) For AAN’s earlier dispatches on the ‘Andar Uprising’, see:
(2) A more detailed account of the battle is as follows:
On 17 October, a Taleban suicide bomber blew his vehicle up near the district compound, leaving the way open for about 30 Taleban to attack on foot; their numbers swelled to 300 the following day. There was a three-day siege, with the district headquarters defended, District governor Muhammad Qassim Disiwal told AAN, by about 60 ANP and ANA. The Taleban blocked the three main routes which the government could have used for sending reinforcements. Two of the roads were mostly passing through Taleban-controlled areas that are located in the northern and western parts of Mirai, while on the third road which connects Mirai to the closest ANA base, in Chardiwal, and then onto Paktika to the east, Taleban fighters set up ambushes near Sultanbagh town, as well as near Ghazni city in the Urzu area. The siege was only broken by ANSF reinforcements, including ANA commandos and intensive air strikes by US forces, starting from late night on the evening on 19 October, 2017. According to villagers near the district centre, the Taleban started to retreat after suffering casualties from the airstrikes.
As well as the seventy or so people killed or wounded, Andar district centre’s infrastructure was left badly destroyed in the latest Taleban onslaught. A newer concrete building in the administration withstood the explosion, but a second, older, mud-built building was left mostly destroyed. Six shops in the Mirai Bazaar also caught fire because of Taleban shelling and burned to the ground. IEDs laid by the Taleban on the road to Chardiwal left the main asphalted road to Paktika damaged in three places near Salam Gudali village. The ANP who fell back to defend the district centre never re-took those checkpoints and the government lost some of the last villages it had controlled.
(3) Our best estimate of the casualties is 46 killed and 31 injured. Conflicting information has been provided by Afghan officials about the number of ANSF, Taleban and civilian casualties, ranging between 20 and 60 fatalities. The father of two ANP who were killed was allowed to enter the district building after sunrise on 17 October to look for their bodies and was reported as saying he saw the bodies of more than a dozen security forces lying on the ground. Officials in Ghazni told a BBC Pashto reporter that 30 people including 25 security forces, mostly Afghan National Police (ANP), and five civilians had been killed. However, a Provincial Council member Abdul Jami Jami told the BBC that 35 security forces were killed. Meanwhile, district governor Disiwal told AAN that ten security forces were killed, another 24 were wounded, ten of them with fatal injuries.
As for the Taleban, local businessmen in the district told AAN that the airstrike had killed 16 of their fighters. However, government officials claimed that more than 90 Taleban fighters were killed, while breaking the siege and in the air bombing that happened after the siege. Taleban spokesman Zabihullah Mujahid told Pajhwok that only seven of their fighters had been killed and as many wounded in the three-day siege.
(4) Asadullah Khaled, who is from Ghazni, had been the Minister of Tribal and Border Affairs when the uprising started. In September 2012, he became director of the National Directorate of Security (NDS). He enjoyed a pre-2001 working relationship with the CIA, a close relationship subsequently also with the US military and the Karzais, and a well-deserved reputation as a torturer (read his biography here.)
(5) Matthew P Dearing is now director of the South and Central Asia Security Studies Programme at the College of International Security Affairs. His analysis of Andar’s ALP and uprising force was published in 2017 by the journal, Small Wars and Insurgencies, “A double-edged sword: the people’s uprising in Ghazni, Afghanistan,” (28:3, 576-608, 10.1080/09592318.2017.1307611) (unfortunately, it cannot be read for free).
(6) Responsibility for security was transferred from the NATO ISAF mission to the ANSF in Andar in the fourth tranche of Transition (Enteqal), announced on 31 December 2012. The fifth and final tranche of Transition was announced on 18 June 2013. This NATO map shows when districts and provinces were transitioned.)
(7) A local businessman named some of those present: Haji Mirza a tribal elder from Qadamkhel, another elder Muhammad Raza from Mullah Muhammad Gudali, a religious figure Mawlawi Ibrahim from Mir Hazar, Mir Ahmad from Ghundi, and Nur ul-Haq Akhundzada from Gandaher (Mir Ahmad and Nur ul-Haq Akhundzada were later killed by Taleban). The only man at the ceremony not to join the ALP, he said, was Haji Wazir from Akal. Another person there, 33 year old Baz Muhammad Khaksar, had been a member of the uprising groupfrom the first. He named some other elders whom he said also participated in the gathering: commander Fatah from Alamkhel, Haji Zahir from Ghundi and 85 year old Haji Amin from Mahkam. All three were later killed by the Taleban, Fatah and Haji Zahir in their houses and Haji Amin outside his village mosque one evening in the month of Ramadan. Baz Muhammad said all three had been prior sympathizers with the uprisersand supported their ‘upgrade’ to ALP.
(8) The estimate was from MP Khial Hussaini, reported here. It counts only the Andar dead and includes combatants. It excludes police, army, Taleban or Hezb-e Islami from other districts who were killed. UNAMA documented 45 civilian casualties (deaths and injuries) in 2012, in Andar as well as 20 ALP and uprising combatants.
(9) The Taleban also carried out a number of ‘preemptive’ assassinations in 2012 of former Qari Baba sub-commanders who could have been persuaded to lead the new pro-government militias. For example, they killed Haji Muhammad from Bagha village and Ashraf Khan from Sher Qala village. Both men were previously with Qari Baba.
(10) UNAMA wrote:
In 2014, the ALP Monitoring and Investigations section of the ALP Directorate investigated 68 accounts of ALP-related human rights violations, including murder, extortion and mistreatment, arbitrary detention, illegal search operations, extortion of “taxes” and forced eviction. Officials in the ALP Directorate reported that the investigations led to 64 arrests and four convictions – the highest number of ALP convictions recorded. UNAMA also noted an improvement in the Ministry of Interior’s case tracking through its new documentation and reporting on the number of prosecutions and convictions resulting from ALP investigations.
(11) The Uprisers turned predatory, Dearing wrote, but ALP “could be considered a relatively protective organisation.” He thinks command and control of the ALP largely worked: “When patrons and the community engaged in complementary governance over the paramilitary group, in this case through the ALP, paramilitary behavior was protection of the civilian population. However, when patrons and communities failed to provide complementary governance, as the case of the remaining Uprising force after ALP institutionalization, the paramilitaries engaged in predation of the local population.”
(12) The Ministry of Interior official said that 32 Shinwari were present in Andar (out of a tashkil of 50). According to current regulations, as the ALP is a defensive force, checkpoints should be no more than one kilometre away from local policemen’s villages. ALP should not be deployed away except under the express orders of the Provincial Police Chief (this happened during the defence of Kunduz in 2015, and Lashkargah in 2016, for example). Deployment to another province and not even a neighbouring one and for routine, rather than emergency duty is far beyond ALP procedures.
(13) Baz Muhammad, his brother told AAN, he joined the ANP in February 2018 and was deployed to Gilan district where he was the commander of four checkposts. On 1 April 2018, an informant asked him to come to a village near his security post, but when Baz Muhammad arrived, Taleban fighters were waiting. They tried to kidnap him, but he fought back, said his brother and, along with two other ANP, was killed.
Two new reports have found that despite improvements in some sectors, aid delivery in Afghanistan is still largely ineffective and poverty has risen. A joint Oxfam and Swedish Committee for Afghanistan (SCA) report on aid effectiveness reveals that while development aid has decreased, donor support continues to be fragmented and aid dependency remains high. Meanwhile, Afghanistan’s new Living Conditions Survey shows that poverty is more widespread today than it was immediately after the fall of the Taleban regime. AAN’s Jelena Bjelica and Thomas Ruttig summarise, contextualise and analyse the reports’ findings.
A new joint report by Oxfam and the Swedish Committee for Afghanistan (SCA) entitled “Aid Effectiveness in Afghanistan” (1) assessed efforts by both donors and the Afghan government to align with criteria provided by the 2005 Paris Declaration on aid effectiveness and builds on a 2008 Oxfam report on aid effectiveness. (2) The report – which was informed by a comprehensive review of the ministries’ and other government institutions’ primary data as well as interviews with key-informants among government, donors, development partners and civil society – comes to the conclusion that the continuing “fragmentation” of aid provided in many sectors still “leads to ineffective aid.” The report states that:
There are over 30 different international donors disbursing aid in Afghanistan, each with their own agenda and aid agreement with the government, and effective donor coordination and harmonisation is not a practice adopted universally […] Yet there are still major issues of fragmentation, with donors bypassing government systems in multiple areas of the development sector, and it is this fragmentation that leads to ineffective aid.
The 2008 Oxfam study had already uncovered worrying facts and figures about how ineffectively aid in Afghanistan had been disbursed and spent in the period from 2001 to 2008. The report revealed that over two-thirds of the aid – then around 20 billion USD in total – bypassed the Afghan government. Donors justified this practice by pointing to the endemically corrupt Afghan system, arguing that channelling money through government and non-governmental organisations guaranteed better financial accountability. This argument was bolstered by the findings in the report that the Afghan government did not know how the remaining third of the aid, which was around five billion USD, had been spent, due to a lack of internal communication and coordination.
The 2008 report also found that “over half of aid is tied, requiring the procurement of donor-country goods and services.” Furthermore, it pointed out that large sums of development money was siphoned off, particularly through reconstruction contracts “for international and Afghan contractor companies,” where profit margins were “often 20 per cent and can be as high as 50 per cent.” It also criticised the exaggerated salaries of “most full time, expatriate consultants, working in private consulting companies, [who] cost 250,000–500,000 USD a year.” It estimated that 40 per cent of aid since 2001 – around six billion USD – had returned to donor countries in corporate profits and consultant salaries, effectively turning aid into donor countries’ export promotion. A World Bank report published in the same year stated that spending “on” Afghanistan did not equal spending “in” Afghanistan, and that “only 38 cents of every aid dollar spent in Afghanistan actually reaches the economy through direct salary payments, household transfers, or purchase of local goods and services.” (More in this AAN analysis)
What does the new Oxfam/SCA report say in detail?
Around 66 per cent of Afghanistan’s budget in the financial year of 1396 (March 2017-February 2018) was funded through international donor support, according to the Oxfam/SCA report. Only 33 per cent came from domestic revenues, even though revenue has tripled over the last ten years, from around USD 750 million in 2008/09 to USD 2.5 billion in 2017/18. These figures reflect a continuing high level of aid dependency. (3) Additionally, the report said (with reference to a 2016 Afghan government update through the Self-Reliance Through Mutual Accountability Framework (SMAF) that 59 per cent of development aid provided by the international community had gone through the government’s core budget that year. But this percentage fluctuates on a yearly basis: in 2015, for example, of 3.73 billion USD disbursed, around 40 per cent was provided through the budget (see here).
The funds Afghanistan receives are mainly channelled through donor-run projects as well as trust funds, which are usually designed as multi-year endeavours and managed by donors. For example, for a major aid investment such as the Afghan Reconstruction Trust Fund (ARTF), although it is a fully ‘on-budget’ programme, programmatic decisions are still made by the World Bank. This is strongly influenced by donors’ investment choices and restricts governmentownership of the ARTF, as the Swedish government’s development agency SIDA reported in a 2015 evaluation (http://www.sida.se/contentassets/72dc94b2318644e9b70242b037660cfd/7a6d0a72-27e6-4bad-b17d-a44c1028ae45.pdf). This approach to budgeting, however, has resulted in the ‘over-aiding’ of the country by allocating more funds than needed for the multi-year cycles. This was also highlighted by the 2016 Afghan Ministry of Finance annual performance assessment report.
The fragmentation of aid is reflected by the fact that funds for the 6.659 billion USD Afghan government budget for 2017/8 were provided by over 30 different international donors. (4) The Oxfam/SCA report stated that each donor maintains its own agenda based on separate aid agreements with the government and highlighted that donors mainly choose to fund areas that appeal to their constituents back home:
With the government largely only able to influence where the money goes, donors are free to fund areas that are appealing to their constituents back home. Therefore, when it comes to the National Priority Programs, where the government has a clear plan for what they would like funded, donors have been known to compete for the most attractive projects, with other development areas neglected if they do not appeal.
In conclusion to its extensive analysis about whether the Afghan government and the international community have met their aid effectiveness commitments pledged in the past four international conferences on Afghanistan (Kabul 2010, Tokyo 2012, London 2014, Brussels 2016), the report reiterated the same finding: development support continues to be fragmented. However, it commended the efforts made by the Afghan government and the donor community to better align and coordinate aid, with the development of national priority programmes (NPPs) and the strengthening of the Joint Coordination Monitoring Board (JCMB), and noticed that in some sectors, however, donors do coordinate well. Mechanisms such as the Afghan Reconstruction Trust Fund (ARTF) and the Afghan Infrastructure Trust Fund (AITF) are given as examples of this “harmonised approach.” SIGAR’s new report on the ARTF, however, disagreed with this Oxfam/SCA finding. According to SIGAR, the World Bank, which manages the fund, “restricts donor and public access to how it monitors and accounts for ARTF funding, leaving donors, and their taxpayers, without important information necessary to understand the activities they fund.” The SIGAR report also states, “the Bank is not following its own policy to provide donors and the public access to ARTF records that should be publicly available.”The World Bank press release in response to the SIGAR report called “most of the findings […] somewhat anecdotal.” Nevertheless, both the SIGAR and Oxfam/SCA report, while they contradict each other on this particular topic, point to poor management, coordination and monitoring of development aid money in Afghanistan. While the Oxfam/SCA report notes this is the case in general for all development aid provided, the SIGAR report states the same for a specific trust fund. Both reports also note that the lack of transparency further exacerbates the unaccountability of development aid funding in the country.
According to the Oxfam/SCA report, the area that has received most financial support is social infrastructure and services, with over 14 billion USD from 2011 to 2015. This is followed by economic infrastructure and services (4 billion USD), humanitarian aid (2 billion USD), and production sector support (1.6 billion USD). It further offers a detailed overview and effectiveness assessment of the projects funded in each of the above-mentioned sectors.
It also found that in 2014 the majority (58.7 per cent) of donor financed ‘off-budget’ projects were below one million USD, and nearly a third between one and ten million USD. “By contrast,” the report stated, “more than a quarter of ‘on-budget’ projects in 2014 were in the 10-50 million USD category and only 21.7 per cent of on-budget projects were less than 1 million USD.” This, according to the report, means that donor support is more fragmented when provided off-budget:
The implementation of a large number of small projects, involving a large number of implementing agencies, despite existing coordination mechanisms, can lead to increased transaction costs for both donor agencies and the Government. Donors may decide to deliver aid this way to reduce their reputational risk, however it can actually increase their fiduciary risk with more resources needed to keep an eye on multiple projects, and eventually increase the long-term development risk as government ownership of this type of development approach remains limited.
It is also evident from the now publicly available and searchable Afghan annual budgets, that the government annual budgets are structured around projects (see here)and not, for example, on sub-national units like provinces or districts.
Additionally, the report states, “a decreased appetite for risk, due to the deteriorating security situation,” the report said, “impacts on their ability to coordinate […] on their ability to share information and agree to common efforts for development advocacy and planning.” This inevitably impacts national development priorities as set out by the government, and often means donors are unable to fund projects in large parts of the country. A development worker who wished to remain unnamed, told AAN that around 40 per cent of the country was not privy to donor funds. This further widens the gap between rich and poor, increasing inequality in society, as mentioned above. A lack of donor coordination and an uneven distribution of donor funds, as well as favouritism of certain projects have contributed to an increase in anti-government sentiments in some parts of the country.
The report does not adequately address the question of the extent to which aid spent by the Afghan government and the international community has been effective (or spent on intended purposes). In the report’s analysis of projects, it is unclear whether they were funded by the government or directly by donors.
What budgeting has to do with aid effectiveness
According to financial experts, the aid channelled through donor-run projects and trust funds almost never uses policy-based budgeting techniques (see here; see also the explanation about ARTF, above), ie the funding associated with these projects lacks the precise identification of public policy objectives, delineation of the means and resources for accomplishing them and an accurate assessment of accomplishments. The funds channelled through projects are linked to “results targets” and “log frames,” the set of qualitative and quantitative measurements that are set in project management paradigms and not budgeting and fiscal accountability paradigms, which should be the case when it comes to state-level budgeting. This, in practice, translates into what the 2008 Oxfam report described as:
[…] A large proportion of aid has been prescriptive and supply-driven, rather than indigenous and responding to Afghan needs. [… Aid] has tended to reflect expectations in donor countries, and what western electorates would consider reconstruction and development achievements, rather than what Afghan communities want and need. Projects have too often sought to impose a preconceived idea of progress, rather than nurture, support and expand capabilities, according to Afghan preferences.
The 2008 Oxfam report highlighted that,according to the 2006 Paris Declaration Survey, only 52 per cent of development aid allocated to Afghanistan during the early years of the intervention had been disbursed “in agreement with the government.” This estimate included all funds provided – both to the core budget and for projects where there was a signed agreement or memorandum of understanding with a ministry or government agency. In the period from 2012 to 2014, according to the Afghan government donor coordination report quoted in the 2018 Oxfam/SCA report, only 4.4 billion USD of development assistance were considered to be aligned with national priority programmes; the remaining 8.4 billion USD were considered counter to estimated needs and priorities. (See also the 2016 Ministry of Finance annual performance assessment).
Over the years, this approach also increased incentives to so-called auction-based budgeting (see here). This, in practice, means that annual budgets in Afghanistan are designed in an auction-like process. This happened most recently in 2017 when the Afghan annual budget was only passed by parliament after the Ministry of Finance agreed to allow each MP to have ‘their’ projects included in the budget (see AAN analysis here). (5) The process of auction-based budgeting, nevertheless, is where the largest incidents of misallocation, mismanagement, and outright corruption usually take place, experts say. The amount of rents, experts estimate, that can be lost through the auction process can be as high as 20 to 40 percent of a budget.
Many reports on Afghanistan reflect exactly this: that a considerable amount of donor funding to Afghanistan has been misappropriated through corruption or misallocation, despite both donor and government efforts to create accountability and transparency mechanisms (see here). A 2016 report entitled “Corruption in conflict: Lessons for the U.S. experience in Afghanistan” from the US government’s Special Inspector General for Afghanistan Reconstruction (SIGAR), for example, strongly criticised its own government for pouring billions of dollars into Afghanistan with so little oversight that it fuelled a culture of “rampant corruption” and undermined the US mission (see also here and here). (6)
Dropping aid levels
The Oxfam/SCA report also provides an overview and analysis of development and humanitarian aid to Afghanistan during the period 2010 to 2015 that equalled 34.3 billion USD. This is part of a total of development aid assistance, between 2001 and 2016, of over 71 billion US dollars in commitments – and slightly over 61 billion US dollars in actual disbursements, according to the Organisation for Economic Co-operation and Development (OECD) (see dataset here). The main donors were the US, UK, EU, Japan, Germany, the Nordic countries, Australia and multilateral institutions such as the World Bank and the Asian Development Bank.
Kabul-based financial experts who prefered to remain anonymous estimate that the country received more than double that amount in military support over the same period; the exact figure is not in the public domain. (7) Recent figures on US spending alone seem to confirm this ratio. According to a 2018 report entitled “Afghanistan: Background and U.S. Policy In Brief”:
… Congress has appropriated more than $126 billion in aid for Afghanistan since [financial year] 2002, with about 63% for security and 28% for development (and the remainder for civilian operations, mostly budgetary assistance, and humanitarian aid).
But aid levels to Afghanistan have fluctuated since 2001. After years of a slow rise, the aid total peaked in 2011, only to decrease significantly afterwards (see here). According to the Oxfam/SCA report, Afghanistan received 6.867 billion USD in 2011, compared to 4.239 billion in 2015(see figure 1 below). The World Bank estimated a decline from an annual average of 12.5 billion USD between 2009 and 2012 to around 8.8 billion USD in 2015. As indicated in this dispatch, records on development aid for Afghanistan vary from source to source.
Figure 1: Development aid disbursed by international community from 2010-2015 (World Bank). Table made by Oxfam/SCA report.
The Oxfam/SCA report also pointed out discrepancies in the reported amounts of development aid for Afghanistan:
In its report, the Ministry of Finance states that Afghanistan received a total of USD 12.9 billion in development assistance in the period 2012-2014, which is less than a figure of 16.8 USD billion reported by the World Bank for the same period. […] The challenge to obtain accurate data on how much development aid has been received from 2010-2106 points to a lack of transparency and coordination within the aid sector of Afghanistan where clear financial data is not readily available, and agreed upon.
The report, in addition, highlights that aid disbursements have been far lower than the pledges. But this is hardly news, as the same finding came out of the 2008 Oxfam report, which said that “donors are far quicker to make promises than to report on disbursements and shortfalls.” In the period 2002 to 2008, out of 25 billion USD committed, Afghanistan actually received less than two-thirds, or around 15 billion USD for development and reconstruction, the 2008 report found. (In the same period Afghanistan received an additional 5 billion USD for the government.)
The new joint Oxfam/SCA report, however, says that there has been some improvement in the ratio between commitment and disbursement. According to the Afghan Government’s Central Statistics Organisation (AGCSO) data, the report noted “the international community has been improving since 2009/10 when only 31 per cent of pledged funding was disbursed, to 2016/17 when 71 per cent of what it had pledged was disbursed.” This point was also reiterated in an earlier International Crisis Group report, which said that between 2001 and 2011, 57 billion dollars of aid money had been spent in Afghanistan against 90 billion pledged (see this AAN analysis).
Changed aid paradigm: Development now for development’s sake?
Another factor that has influenced aid fluctuations is that ten years after the 2008 Oxfam report, the donor landscape in Afghanistan as well as the dominant development paradigm has changed. Some of the major aid actors, ie the international military, which distributed the lion’s share of development and humanitarian aid through its Provincial Reconstruction Teams (PRTs), have almost completely disappeared. The PRT Commanders Emergency Response Program (CERP) set up by the US, for example, spent about 1.5 billion USD alone between 2004 and 2011. But this high spending contributed little to the desired outcome of stabilisation, which had been the dominant development paradigm up until then.
As noted in 2009 by Andrew Wilder, “given the centrality to the counterinsurgency (COIN) strategy of the assumption that aid is an important stabilization tool, and the billions of development dollars allocated based on this assumption, there is surprisingly limited evidence from Afghanistan that supports it.” Wilder also noted that despite the amount of money spent, “Afghan perceptions of aid and aid actors had been overwhelmingly negative” (see also this Tufts University study on PRT in Helmand here). In contrast to the 2008 Oxfam recommendation to increase development aid to Afghanistan at the expense of military support, donor support to development and humanitarian needs has decreased over the years.
Aid dependency still high
Afghanistan’s aid dependency rate, meanwhile, remained extremely high. Between 2005 and 2011, it averaged 76 per cent of the country’s GDP (see here). This means that over that period Afghanistan’s aid dependency was more than three times higher than the rule of thumb proposed for optimal aid levels for a functioning, sovereign state by the Washington-based Institute for State Effectiveness, at around 20 per cent of GDP (see here). According to the World Bank, this rate decreased to 45 per cent in 2017, still more than double the recommended rate.
In short, the country has been swamped with money since 2001, which has driven its heavy dependence on development aid to new levels. Between 1977 (the last full year of the Republic) and 1988 (the last full year of Soviet troop presence), foreign (mainly Soviet) aid roughly covered between one fourth and one third of state expenditures, according to a 2007 Canadian study (see here).
Rising poverty as a key outcome?
Another trend becomes apparent when reading the Oxfam/SCA report and the Afghanistan Living Conditions Survey released on 2 May 2018 by the Central Statistics Organization (CSO) in cooperation with the Afghan Ministry of Economy and the World Bank (see here): that the large amounts of aid disbursed since 2001 have not reduced poverty in Afghanistan. This is nothing but a result of continuing ineffectiveness. While the rate of Afghans living in poverty was lowered from the 2003 World Bank’s baseline of 51.4 per cent of the population in 2003 – ie during the immediate post-Taleban era, when aid flows were still moderate – (see here) to around 34 per cent in 2007/08, these gains were subsequently lost again. The paradigm shift around 2011 – which should have been in the right direction, making development the paradigm instead of stabilisation – was accompanied by a slump in aid money alongside the gradual withdrawal of foreign troops, the transition (‘handover of responsibility’) and the Afghan government’s paradigm of going from transition to ‘transformation’. This seems to have resulted in a higher poverty rate, once again. It increased to around 38 per cent in 2011/12, which was, until recently, the last official CSO figure (see here). (8) It is now 54.5 per cent. In fact, this is even higher than the 2003 figure in the immediate post-Taleban period. During the Taleban regime, international aid was as low as 564 million USD for the entire period between 1995 and 2000, according to the OECD data set already quoted above. This, combined with the non-policies of the Taleban, led to widespread impoverishment.
In 2015, the Afghan Ministry of Economy explained the lack of poverty reduction in spite of growing development assistance in its Afghanistan Poverty Status Update by an increase in inequality, measured by the Gini index. (9) According to the document, this index in Afghanistan increased from 29.7 per cent in 2007/08 to 31.6 in 2011/12, when the country received the highest amount of development aid (see figures 2 and 3). It stated, “Had Afghanistan’s economic growth been more widely shared, poverty could have declined by as much as 4.4 percentage points.” That the Gini index had decreased by 2016/17 (despite an increase in poverty) does not mean, however, that growth was being shared more widely or more equally. It is a statistical phenomenon resulting from an almost 20 per cent increase in the poverty rate.
Figure 2: Data downloaded from the Central Statistics Organization (CSO) dashboard. Table made by AAN.
Figure 3: Data downloaded from the Central Statistics Organization (CSO) dashboard. Table made by AAN.
A possible way forward
It is expected that at the Geneva donor conference in November this year, the 2016 Brussels conference commitments will go through a mid-term review. It also means that a new set of indicators and targets could be designed to measure Afghanistan’s progress towards self-reliance. At the same time the question remains open as to who will measure how donors align themselves with the 2005 Paris Declaration and how they will deliver on their financial promises. The non-governmental Oxfam/SCA report may be a good example of an exercise in self-scrutiny, but there is still a lot to be done by other actors, too, particularly among governments. Part of the problem concerns Afghanistan’s chronic corruption and points to a lack of statesmanship among the country’s political élite. They see power as a way of accessing ‘rent-seeking’ while proving unable to address the large and rising portion of their fellow Afghans languishing in poverty. This requires a behavioural change on their part. But an equally radical change is probably required by donors, who need to review their way of doing business in Afghanistan, in line with some of the recommendations made by the recent reports on aid effectiveness.
Edited by Sari Kouvo
(1) The Paris criteria (from the Oxfam/SCA report):
(2) Oxfam and SCA had the support of Catholic Agency for Overseas Development (CAFOD) for this report. The report was prepared by ATR Consulting in Kabul, Afghanistan, the views and recommendations presented in the report are those of ATR, and do not necessarily reflect those held by SCA, Oxfam or CAFOD.
(3) This is also evident from the budgets for the last seven years, which are now available to the public on the Ministry of Finance’s website (see also AAN’s analysis on the Afghan budget for 2018/19). Publicly available and in searchable formats (PDFs and excel) budgets for the last seven years, are the latest attempts of the current Afghan government to make its public finances accessible and transparent and in line with the best international open government practices.
(4) According tothe Oxfam/SCA report the budget was split into two categories: 62 per cent for the operating budget and 38 per cent for the development budget. Donors gave money to both budgets, 1.9 billion USD for the operating budget and 2.214 billion USD for the development budget. In the development budget, donors provide both discretionary (493 million USD) and non-discretionary (1.674 billion USD) funding (see here).
(5) A recent Afghan media report explains how this works in practice, quoting an Islamic scholar and a civil society representative speaking at an election-related press conference in early May 2018:
[Mawlawi] Rahmatullah Andar, an Islamic scholar, told the press conference that […] “The government approved 75 development projects for Ghazni [province] this fiscal year and reserved only 14 to as many districts while the rest of the other projects are given to four districts which have representatives in the Wolesi Jirga.”
A civil society activist from Ghazni, Inayatullah Naseri, also said only four people from as many districts out of Ghazi’s 18 [district centres] represented the whole province in the parliament.
(6) A leading Afghan anti-corruption civil society organisation, Integrity Watch Afghanistan’s (IWA) 2016 national survey on perceptions and experiences of corruption, found that more than 70 per cent of Afghans thought that corruption in 2016 was worse than it was in 2014 when a similar survey was conducted.
(7) According to CSIS figures (quoted in this AAN analysis), the US ratio between military and civilian spending between 2001 and 2012 in Afghanistan was 16:1.
(8) There are more recent World Bank figures (see here), putting the 2011/12 poverty rate at 36 per cent and giving one of 39 per cent for 2013/14 – deviating somewhat surprisingly from the CSO figures, despite the fact that the World Bank cooperated with the CSO on the figures in the Afghanistan Living Conditions Survey, formerly known as the National Risk Vulnerability Assessment.
(9) The Gini index or Gini coefficient is a statistical measure of distribution developed by the Italian statistician Corrado Gini in 1912. It is often used, including by the UN, as a gauge of economic inequality, measuring income distribution or, less commonly, wealth distribution among a population (see here).