International crisis instant coffee: the Bangladesh–Myanmar border region

So many people wanted to cross the river [from Myanmar into Bangladesh]. There were Burmese helicopters above us, shooting down at us. Then they started throwing grenades. Everyone started running towards the river. People with money were able to cross first. There was no fixed rate [charged by the boatmen]. Of the people without money they simply took the nose jewellery or whatever they had left.1

the crisis in 2017. Rohingya are not the only people who cross the border to escape violence in Myanmar; daily consumption goods, drugs and weapons are also smuggled across. For instance, the yaba that is produced in Myanmar's Shan State is smuggled in large quantities, catering to an extensive consumer base in Bangladesh (see also Figure 3.4).

BACKGROUND: A BRIEF HISTORY OF ARMED CONFLICT IN RAKHINE
The recent wave of violence against the Rohingya in Myanmar has to be understood against the backdrop of a long history of marginalisation and attempts to drive the community out of the country in order to create a national identity centred on Buddhism. The community that is commonly described as 'Rohingya' is a predominantly Muslim group of people from Arakan, what is today Myanmar's Rakhine State. The earliest records of the term 'Rohingya' exist from the late eighteenth century, but a Muslim population already lived in the area at the time of the independent Kingdom of Mrauk U (1429-1785) (Poling, 2014 Myanmar government does not use the term 'Rohingya', but calls the community 'Bengali' and views them as indelibly foreign. Burma issued National Registration Cards to all residents whose families had been in the country's territory for at least two generations in 1952, including many Rohingya (Wade, 2017). However, after General Ne Win took power in 1962, they were gradually stripped of their rights and declared to be illegal Bengali immigrants (ibid.). Their National Registration credentials were replaced with Foreign Registration Certificates in 1974. In 1978, the state displaced 200,000 Rohingya to Bangladesh during a military campaign. In 1982, as part of a new Citizenship Act, all Rohingya living in Myanmar became de facto stateless as they were not considered to be one of the 135 'national races' that now defined citizenship (ibid.).
A number of Rohingya political groups with armed wings evolved in the late 1970s and early 1980s seeking to fight for the community's rights (Brennan and O'Hara, 2015). The best-known group was the Rohingya Solidarity Organisation (RSO) (ibid.). RSO was supported by Jamaat-e-Islami Bangladesh (BJeI), a major political party that had pursued the establishment of Islamic principles and code of law in Bangladesh until the country outlawed it in 2003(Fair et al., 2017. With this backing, RSO was able to maintain bases on the Bangladeshi side of the border and train its fighters there. The Arakan Rohingya Islamic Front (ARIF) had splintered from the RSO in 1986, but they reunited in 1999 and became the Arakan Rohingya National Organisation (ARNO) (see ARNO, 2006;ICG, 2016, p.4). The consolidated armed wing was named the Rohingya National Army (RNA).
Meanwhile the Burmese government established a complex social engineering project, called Na Ta La, in 1988(Wade, 2015. It aimed at weakening the Muslim population of Rakhine State by encouraging Buddhists from other parts of the country to migrate there (ibid.). In the mid-1990s inmates from prisons were offered early release when they agreed to resettle in Rakhine (ibid). In the 1990s the government also began to limit the movements of Rohingya in the country, making them obtain special permits in order to leave their townships (Wade, 2017). Efforts to settle more Buddhists in northern Rakhine (Figure 4.2) continued (Rahman, 2013). After 9/11, the Myanmar government embraced the US discourse of fighting a global 'War on Terror' (Allchin, 2019, p.131). It actively reached out to the Americans, sharing 'intelligence' with the US embassy in Yangon and attempting to link the Rohingya resistance to global terrorism by claiming that ARNO was in contact with Osama bin Laden and had its fighters trained by Al Qaeda in Libya and Afghanistan (ibid.). However, according to Allchin, these claims were unfounded. According to his analysis, ARNO at this point was nothing more than a 'rag-tag group of aspirant organisations, some of which had perhaps a limited stock of arms' (ibid., p.132).
In the early 1990s around 230,000 Rohingya refugees left Myanmar for Bangladesh, living in camps close to Cox's Bazar. Most of them were repatriated to Myanmar in the following years, many of them forcibly (Refugees International, 2005 Figure 4.2). The Myanmar government reported that nine police officers were killed and 62 firearms were stolen (ibid.). In August 2017, Harakah al-Yaqin was renamed 'Arakan Rohingya Salvation Army' (ARSA), shifting its communicated ideology from a religious base to an ethno-national one. The Myanmar government considers ARSA to be 'terrorists'. Another growing insurgency group is active in Rakhine State and is fighting against the Tatmadaw: the Arakan Army (AA). The AA was established in 2009 in Kachin State and was trained by the KIA in the early years. It views itself as representing the indigenous Buddhist Rakhine people and is fighting for independence and self-determination of Rakhine State. Drawing on its links with the KIA, the members of the AA not only receive a much more professional training than members of ARSA and are considerably better equipped with weapons, but the AA is also much larger; media reporting estimates that AA has around 7,000 fighters (The Irrawaddy, 2019a). Furthermore, the AA is suspected to be active in Bangladesh's Chittagong Hill Tracts (CHT) north of Cox's Bazar, which enables the group to operate across the border.

The August 2017 Escalation of Violence
On 25 August 2017, ARSA supporters, most of them untrained and only armed with machetes and sticks, conducted simultaneous attacks on an estimated 30 small BGP and army outposts (ICG, 2017b, p.6). According to the Myanmar government, 14 soldiers and 371 ARSA members were killed in the course of the attacks (ibid.). Three days later Ata Ullah directed ARSA's fighters to burn down Rakhine Buddhist villages. Most disregarded the order but three villages were attacked (ibid., pp.6-7). In one village in Maungdaw Township, ARSA members allegedly killed a large group of Hindus (ibid., p.7). While survivors first blamed Buddhist nationalists, Myanmar media later reported that the massacre was conducted by ARSA (ibid.). This case gained international attention through a subsequent Amnesty International (2018a) investigation, which reported that ARSA was responsible for the killings.
Following the attacks on 25 August, the Myanmar Tatmadaw, the BGP, the police, and groups of vigilantes pursued a large-scale campaign of violence against Rohingya in Rakhine State that lasted until 24 September 2017. The Independent International Fact-Finding Mission on Myanmar states in its final report that 'the operations were designed to instil immediate terror, with people woken by intense rapid weapons  Many interviewees shared their experience of 25 August and the following days in their interviews with me. For example, a young man reported: On 25 August the problems started. The military came to our village at night. They stayed for about two hours, killing three children. Everyone else escaped from the village. We were worried that they would come back. The next day some people decided to leave for Bangladesh. I and some others decided to stay. I stayed with relatives close to my village. Two days later the military did come back. They burned half of the village. I saw it myself. They stayed for maybe five hours. I left a few days later. It was raining a lot. When I crossed the border river by boat the BGP and the military shot at the boats. Some boats capsized and many people drowned. 4 He shared with me the photos (Figure 4.4) he took of his house after it was burned. We saw what was happening from the other side of the river. Some people started fleeing. Others were surrounded in the village. They separated eight women from a group, probably to rape them, and shot everyone else. In another group, they separated 20 to 30 women and took them to houses, taking turns raping them. Then they burned the houses with the women inside. I saw 12 groups of dead bodies. They put petrol on them and burned them. It was raining at the time. I hid in the hills for another day and then came to Bangladesh. 6 According to the Fact-Finding Mission's report, 'The few women who survived, and who spoke with the Mission, displayed both serious burn marks and stab wounds, which were consistent with their accounts' (Human Rights Council, 2018, p.182). It is estimated that 750 people were killed in Tula Toli village in August 2017 (ibid., p.183). Many interviewees reported that the Myanmar Tatmadaw had been trying to create divisions between Buddhist Rakhine and Muslim Rohingya in their communities, for instance through the Na Ta La programme, in the course of the previous years. Nonetheless, there were also some reports of Buddhists trying to help the Rohingya in the wake of the violence that occurred in August 2017. For example, an interviewee told me: We used to have a good relationship between Rohingya and Buddhists in our village… Our village leader was Buddhist. When the situation got harsh he called all Rohingya and told us not to worry, he wanted to protect us. But then the military came, threatened him and arrested him. The next day the military came to burn the village. I got woken up by gunshots in the morning… I only took the clothes I was wearing and ran… We later called the village leader and complained, but he said he couldn't help because he got arrested. He tried to help and protect us. He didn't even charge us high amounts of money to protect us as other village leaders did. 7 The Rohingya who fled from Rakhine and escaped across Naf river into nearby Bangladesh joined an existing community of at least 200,000 Rohingya (some estimate up to 500,000) who had arrived in the previous decades. They now live in numerous camps between Cox's Bazar and Teknaf and are confined to this area (Figure 4.5). There are Bangladeshi army and police checkpoints on the road leading to Cox's Bazar, where Rohingya who try to leave the camps were stopped at the time of my research.
While the Tatmadaw and affiliated actors perpetrated much of the atrocities interviewees reported, the civilian government of Myanmar is accused of bulldozing and handing over the land previously owned by the Rohingya to Buddhists. In February 2018, after analysing satellite data, Human Rights Watch (2018)  This claim was also reiterated by some of the interviewed Rohingya in the refugee camps in Bangladesh. Others stated that they 'had never heard of ARSA', 8 'had never seen ARSA members', 9 or that 'ARSA is just an excuse used by the Burmese military to arrest people.' 10 However, still others were supportive of ARSA and viewed it as the only active defence left for the Rohingya. They said that people did not talk about ARSA in order not to create risks for its members, some of whom had allegedly been arrested by Bangladeshi law enforcement or intelligence, or else were too afraid to talk about ARSA as they feared retribution.
Regardless of these competing perspectives, an entity that calls itself ARSA certainly exists. Not only does ARSA have a polished Twitter account, there were also people in the refugee camps who self-identified as ARSA members. Ata Ullah's whereabouts are unknown, with some claiming that he remains in northern Rakhine and others stating that he is in Bangladesh, but most of the rank-and-file members have left Myanmar with the other Rohingya refugees. However, it took a long time to identify them and all of them appeared to be very afraid of being interviewed.
Those ARSA members who did agree to be interviewed describe the main goal of their group as fighting for 'citizenship' 11 or 'general rights'. 12 Some also said they wanted to be able to live according to Islamic Law. 13 They stated that they were not fighting for independence but simply wanted more rights, and the ability to return to their land  15 The details of the training are kept secret. One interviewee explained that he had to promise 'a big vow' to not talk about its details. 16 I suspect that one factor prompting the secrecy surrounding the training is its low quality, given the rudimentary weapons and lack of experience of the trainers, which the interviewees pointed out. One interviewee stated that even during the training sessions they always only use wooden fake weapons, never touching real ones.
According to interviewees, commanders had established the impression that other ARSA groups and ARSA members had guns, and that they would just have to wait for guns to arrive. For instance, one group was told 'your group doesn't have guns, but the groups further up in the north do.' 17 Others were told they would get guns after their training. 18 But then after training they were told 'if you follow Allah everything will be okay. You will get arms if you follow Allah.' 19 Only one of the interviewees had ever used a gun, and he said that four other members of his group had a gun. He said, 'I did have a gun during the attacks in August. But I got a 9mm instead of a rifle.' 20 Similarly, a senior commander told me that some of his subordinates had rifles and pistols. 21 According to the interviewees, the Tiger group had about 200 members of which around 40 had access to guns. They explained that they had only been successful in stealing guns from police stations or military barracks through raids once or twice, which they had carried into battle on 9 October 2016. 22 Generally, the lack of guns was repeatedly brought up as a main concern by ARSA members. Rank-and-file ARSA members said that they had not been particularly violent. They stated that their main responsibility was to watch and warn the women in case of arriving military forces. One interviewee explained: 'When we saw the military, all the women left to another village. Men tried to remain alert, but we didn't fight back.' 23 When I asked them directly about the accusation that ARSA's actions had triggered the Tatmadaw's response, the interviewees claimed that ARSA's mission was resistance against the violence the Rohingya community had been experiencing for a long time. One interviewee stated: 'ARSA was resisting. The genocide did not happen because of our attack.' 24 Another interviewee asked: 'What other option do we have to stand up against the government?' 25 In addition to the lack of equipment and the crack-down of the Tatmadaw on the Rohingya, ARSA also has to deal with the growing influence of the AA. The AA and ARSA share an enemy: the Tatmadaw. Both the indigenous Rakhine Buddhists and the Muslim Rohingya are historically part of the same Arakan Kingdom. The Tatmadaw claim ARSA and the AA are collaborating (e.g. The Irrawaddy, 2019b), but this seems unlikely since the relationship between Rohingya and Rakhine is often strained. Instead, the relationship between the groups is more likely to be one of mutual recognition and non-interference. The AA strongly denies any collaboration and, in March 2019, the deputy leader of the AA, Nyo Tun Aung, stated publicly that the group would even be willing to cooperate with the Tatmadaw to crack down on ARSA (see Mon, 2019).

The Smuggling of Goods
It is not only Rohingya who cross the border from Bangladesh to Myanmar at Cox's Bazar. Whenever I stopped at one of the many small shops in the refugee camps at Cox's Bazar for a coffee break, the coffee was made in Myanmar. This instant coffee is quite easy to identify without even reading the label because of the '3 in 1' nature that is popular in Myanmar, containing coffee, milk and sugar -most likely in reverse order.
I found it intriguing that simple goods such as coffee, which could also be purchased easily in Bangladesh, are smuggled across a border at which many have been shot or drowned. While the border was formally closed at the time of my research, a range of products from Myanmar were available in the Rohingya refugee camps. Smuggling routes for such goods have existed for a long time as the communities on both sides of the border are closely connected, not least because Rohingya have been displaced to Bangladesh repeatedly since the 1970s.
Illicit goods are also traded across the border and have been for a long time, independently of the recent migration of Rohingya refugees. The yaba produced on a large scale in Myanmar's Shan State, as well as other drugs, used to be smuggled across the Naf river into Bangladesh on fishing vessels. There is limited data available on the scale of yaba use in Bangladesh, but it has clearly become an immensely popular drug in the country. It is widely and easily available, not just in Cox's Bazar but across the country, including in Dhaka (see e.g. Thompson, 2017).
According to Bremner (2018), 'boatmen would hide the pills in onions, watermelons, or in plastic bags that they sunk underwater and attached to the gunwale of their boats.' However, with the influx of Rohingya, the refugees have turned out to be a cheaper alternative for the smuggling networks. Some Rohingya refugees brought yaba with them, knowing it to be a valuable currency in Bangladesh. But, more crucially, smuggling networks in Bangladesh now use refugees and ask them to go back to Myanmar and return to Bangladesh with more yaba (ibid.). Desperate to make a living, refugees are willing to take big risks for low payment. From the border area, the drugs are smuggled to the refugee camps. The smuggling networks appear to draw particularly on Rohingya women for this work, as they are less likely to be checked (ibid.). Once in the camps, the drugs can be hidden easily, due to the camps' enormous size and a 7pm curfew for Bangladeshi officials. This scheme puts most of the risk on the vulnerable Rohingya while the powerful businessmen from Bangladesh profit.
According to many interviewees, it is an open secret in Cox's Bazar that one person running much of the criminal activities in the area is a local MP, a member of the ruling Awami League party, and his family. The MP is sometimes referred to as the 'Yaba Godfather'. Even the Bangladeshi Department of Narcotics Control describes him and his family as the key players in control of the yaba 'import' from Myanmar (Haque and Barua, 2014). Following the beginning of a 'war on drugs' in Bangladesh, the MP briefly left the country in June 2018 (Saad and Rahaman, 2018), but he has since returned to Bangladesh. In December 2018, after I had concluded my research, his wife took over his seat in parliament. In February 2019, some of his family members surrendered in a formal ceremony. However, the former MP did not attend. In addition to drugs, guns are smuggled across the border. While on a smaller scale overall, a number of large shipments of guns have been reported. In the past, shipments of weapons from China and Thailand were smuggled on boats to Cox's Bazar or Chittagong, often via Myanmar. From there they were transported on trucks via the Chittagong Hill Tracts, supplying, for instance, insurgency groups in northeast India and western Myanmar (Reddy and Reddy, 2007, p.70). In 2004, ten truckloads of weapons, including sub-machine guns, rocket launchers, grenade launchers and hand grenades, were discovered in Chittagong's port (BEI et al., 2013, p.29; Habib, 2004). In the trial that followed, Motiur Rahman Nizami, the former leader of the Bangladesh Jamaat-e-Islami party, the former home minister Lutfozzaman Babar, and two senior intelligence officers, as well as ten others, were sentenced to death. In 2010, a plan by the insurgency group NSCN-IM, based in northeast India, to ship 1,000 AK rifles and 50 machine guns from the China-Vietnam border to Cox's Bazar was foiled (bdnews24, 2014).
While Chittagong and Cox's Bazar remain prominent entry points for smuggled guns (see e.g. Daily Asian Age, 2018), the area around Cox's Bazar has also evolved into a production hub for weapons. In 2017, an illegal arms factory was discovered in Cox's Bazar (see Dhaka Tribune, 2017a). According to media reports there are around 50 illegal arms manufacturing workshops in the Greater Chittagong area (Dhaka Mirror, 2013). Hence, guns are now also smuggled from Bangladesh to Myanmar. An ARSA member told me that the group also tried to smuggle weapons from Bangladesh into Myanmar, but Bangladeshi law enforcement officers captured the gun runner. 26

The Smuggling of People
At the time of my research, however, what was smuggled the most were people: Rohingya escaping from Myanmar. As Bangladesh opened its border to the Rohingya in 2017, the crossing did not formally qualify as 'smuggling' anymore. Nonetheless, the large number of people who wanted to cross the Naf river (Figure 4.6) from Myanmar to Bangladesh was an opportunity for people who had been active in the people smuggling business before, such as local fishermen.
The Rohingya refugees I interviewed described fluctuating prices for the border crossing, depending on time and geography; around BDT 10,000 (c. US$120) was common, and many had to hand over their jewellery to the boatmen to be taken across. Some refugees stated that the boatmen only took them to the middle of the river and then demanded more money, threatening to let them drown if they did not pay. Conversely, the interviewed Bangladeshi boatmen described themselves as helpful rather than exploitative. They claimed that they simply wanted to help when they saw people stranded at the shore, that they asked for no money but that the refugees insisted, 'we will give you whatever we have'. 27 Rich refugees paid well and poor people did not. It is certainly possible that some boatmen told the truth, but the refugees' statements suggest this was a business opportunity for many.
Transporting refugees is an established business in the area, and people have been fleeing Myanmar, sometimes via Bangladesh, for Thailand and Malaysia for years. Estimates say that more than 60,000 Rohingya were in Malaysia by September 2017 (NST, 2017). The interviewed boatmen described having taken up to 60 people, all a particular boat could hold, three to four times per month out to bigger boats waiting to take them across the Andaman Sea. They described this as a highly organised business run by a Cox's Bazar-based mafia-like group, in which always the Florian Weigand -9781789905205 Downloaded from Elgar Online at 11/05/2020 12:53:18PM via free access same middlemen, who were Bangladeshis from Cox's Bazar, approached them and informed them about the next job. They would be paid BDT 1,000 (c. US$12) per smuggled person. This was the starting point of a risky journey, on which many refugees later crossed through the Thailand-Malaysia border (Figure 4.7). A female interviewee in one of the refugee camps in Bangladesh reported: As mentioned in Chapter 3, the smuggling of refugees can turn into human trafficking, as some of the refugees were later kept in camps in the Thailand-Malaysia border area as hostages until their family members paid for their release. Others were held captive, sometimes raped, tortured and killed, on fishing vessels (Fortify Rights and Suhakam, 2019). In August 2017, there were 300 cases against human traffickers pending at the court in Cox's Bazar. However, all of these cases were against middlemen or boatmen, rather than more senior persons within the networks (see Illius, 2017). A refugee reported what had happened to his brother: In 2012, my brother left for Malaysia. This is when the atrocities in Myanmar started. Many young people got arrested. So he decided to leave. He took a small boat from the shore and there was a bigger boat waiting for them. The Burmese border guards charged everyone another 1 lakh before they allowed them to get on the boat [MMK 100,000; c. US$75]. When they got to Thailand, they got detained by the smugglers. I got a phone call and I could hear my brother crying. He told me that he hadn't had any food in five days and that they would kill him. We managed to collect 5 lakhs [MMK 500,000; The route to Malaysia c. US$365) for them. We gave the money to a shop in Maungdaw. The shop is owned by a Buddhist. But the smuggling business is run by a mixed group of Rohingya and Buddhists. But instead of releasing him, they sold him to another group. He worked for them for six months for free as a mechanic. This is happening to many people: they are kept as hostages. 30 Since the recent influx of Rohingya into Bangladesh, the camps and the borders in the area have been tightly controlled, preventing people from moving further into Bangladesh or leaving for Thailand and Malaysia by sea. While rich refugees were still able to travel to Malaysia using fake documents and air travel, smuggling via boat had virtually come to a standstill at the time of my research in 2017/18. However, the interviewed fishermen stated that they were waiting for the border Florian Weigand -9781789905205 Downloaded from Elgar Online at 11/05/2020 12:53:18PM via free access controls to be relaxed again to resume their activities in support of the migrant smuggling networks. Considering the large number of people in the camps, many of whom already have relatives in Malaysia, large numbers of refugees can be expected to try their luck and to take a boat to Malaysia, once the security measures have been relaxed, and once they feel certain that they have no future in Bangladesh and no chance of returning to Myanmar. Rohingya refugees, including those from previous waves of migration and displacement, are not allowed to leave the camps and work or study in other parts of the region or Bangladesh. However, an organised network provides Bangladeshi identity papers for those who can afford it. Further, at the time of the interviews, a growing number of Rohingya women and children were trafficked out of the camps. An organised criminal network from Cox's Bazar was promising young women and children work, for instance as domestic helpers, as a ruse to force them into sex work (see also BBC, 2018b; Stratford, 2018; UNICEF, 2017). It is unclear whether a single centralised network perpetuates this trafficking or whether there are several criminal groups with different areas of business -such as smuggling of people to Malaysia, providing documents to Rohingya, and human trafficking. Sometimes the recruiters are Bangladeshi, whereas in other cases Rohingya women are used as middle-women to convince other Rohingya, in keeping with common human trafficking approaches across the region.
Bangladeshi army officers admitted that they neither had a mandate nor the means to work successfully against this mafia-like threat. 'If you get into their way, they simply kill you', 31 a soldier stated. They also described a second form of trafficking out of the camps for sexual exploitation. According to an army officer, many Bangladeshi men who do not appear to be part of any network and who often travel from other parts of Bangladesh to the camps take women out with a promise of marriage, intending to sell them or have them as a 'second wife'.

Smuggling and Insurgents
Armed conflict in the Bangladesh-Myanmar border region has created numerous business opportunities for criminal networks. Insurgencies are commonly assumed to be part of or even in charge of these networks. ARSA has been blamed repeatedly, especially by the Myanmar govern-Florian Weigand -9781789905205 Downloaded from Elgar Online at 11/05/2020 12:53:18PM via free access ment (e.g. The Irrawaddy, 2019b), for being involved in the smuggling of drugs (e.g. Dhaka Tribune, 2017b). The group denied such accusations on Twitter, arguing that 'there are other armed groups, dacoit groups [bandits], human trafficking groups, drug trafficking groups and some other groups (…) disguising [themselves] as the members of ARSA with the intention to tarnish the noble image of ARSA in the eyes of the international community' (ARSA, 2018). Indeed, so far no evidence has been found that supports such an accusation (e.g. RFA, 2017). Also my research does not suggest ARSA has the means, capacity and level of organisation to be an important actor in the drug trade, given that they cannot even obtain sufficient weapons to carry out insurgency actions. Furthermore, as a Rohingya insurgency, ARSA is unlikely to be a preferred business partner for the yaba producers in Myanmar's Shan State and, as Allchin (2019, p.137) points out, they lack the patronage networks in Bangladesh that would enable them to operate there. Nonetheless, other non-state armed groups may very well be benefitting from the drug trade. For instance, the AA in Myanmar is often considered an involved actor, and they have networks in northern Myanmar and Bangladesh's CHT, which affords them good access to both the production and sales markets of yaba (see e.g. Allchin, 2019, p.137). Conversely, as a non-state armed group, it is likely that ARSA is involved in weapon smuggling. The insurgencies in Bangladesh, India and Myanmar need arms, which in some cases are smuggled by criminal networks through Cox's Bazar. Cox's Bazar is only one of the hubs for the arms trade. In Myanmar, many groups can more easily source weapons from China or Thailand. However, ARSA, with its much more limited networks and support in the region, has had not much success in its attempts to purchase weapons in Cox's Bazar or elsewhere and smuggle them to Rakhine State. Instead, ARSA obtained most of its weapons from raids on compounds that belong to Myanmar's security forces.
As with drug smuggling, I found little evidence that ARSA is involved in people smuggling, and AA as well as other political non-state armed groups seem unlikely to be involved. In one exception, interviewed ARSA members were still going forth and back between Myanmar and Bangladesh at the time of my research, essentially smuggling themselves. One fighter reported: 'I went [to Myanmar] 14 days ago. But the army noticed. So I had to escape.' 32 But it is not only ARSA members that cross the border like this; others have returned for various reasons, such as checking on what has happened to their property. In another exception, ARSA may have played a limited role in people smuggling that could be described as an unsuccessful humanitarian approach. According to some ARSA members, the group tried to help refugees by renting six boats from fishermen for them to take Rohingya across the river for free. However, this initiative failed as the boatmen nonetheless charged the refugees, and the initiative was therefore abandoned. Other ARSA members explained that in some cases they helped people with cash to pay the boatmen and afford the journey to Bangladesh. In addition, both ARSA members and refugees stated that the group in some cases protected refugees at the shore of the Naf river by forming a protective circle around them, even though they were not armed either. An ARSA member claimed: 'Every day we went out to protect people.' 33 An interviewee who received such help explained: 'ARSA formed a circle around us when the army was shooting into the crowd at the beach. They were not armed and many got killed. But even more civilians got killed nonetheless.' 34 However, many interviewed refugees reported that they had never heard of such incidents.

Smuggling and the State
As I found at other borders, state actors are a considerably more important player in the smuggling economy at the Myanmar-Bangladesh border than non-state armed groups, both with regard to the smuggling of people and goods.
On the Myanmar side of the border in Rakhine State, corruption -in the form of paying bribes to the Tatmadaw -enabled some Rohingya to escape from the violence in some cases, according to interviewees, even during the brutal clearance operation in August 2017. An Amnesty International report found that 'In most cases, only those who were able to pay huge bribes to the security forces were released, leaving those from poor families at greater risk of prolonged detention and further torture' (2018b, p.26). Also the Myanmar coastguard benefitted from the Rohingya refugees. Interviewees reported that even though they were not charged at the time of the mass-exodus in August 2017, the Myanmar coastguard used to charge people a 'fee' allowing them to get on a boat, when they had tried to leave the country in the past. Evidence is similarly scant regarding state authorities on the Bangladeshi side of the border, but what exists suggest they are a noteworthy actor in the smuggling economy. It is not just corrupt officials that look the other way -there appears to be a more direct and systemic involvement. This can be illustrated by what I was told when I asked about the trafficking of Rohingya out of the camps. A Bangladeshi army officer in the camps told me that they were arresting around 50 people a day, whom they caught attempting to smuggle women or children out. 35 This was hardly a surprising number, since every day when I left the camps in the evening, I experienced a significant number of checks by the military. 36 Based on their mandate, the officer explained, the army had to hand over all people they arrested to the local police. Acting on this, I went to the police station, hoping to meet some of the traffickers or, failing that, to gain an understanding of how the police treat them. The responsible high-ranking police officer at the station told me that they only had a very few, isolated cases of people smuggling and human trafficking and that there currently were no people under arrest. Human trafficking, he stated, was not a major concern. 37 It would be premature to accuse the police of corruption, possibly releasing large numbers of arrested people daily. However, considering the vast difference in the accounts of the army and the police this seems to be a likely explanation, even if the army's statement were exaggerated. Evidence also suggests the local Bangladeshi police are not preventing the trafficking of women and children out of the camps. Looking at the drug smuggling activities in the border area illustrates that some local authorities in Bangladesh are part of or closely connected to criminal networks -as not least the case of the local MP shows. If dealers pay a certain weekly fee, these state-linked networks ensure that they and their drug mules remain unharmed. A drug dealer interviewed by a newspaper stated that they had to pay 'up to $120 a week on bribes to create safe passage for their mules' to 'contacts' (Bremner, 2018). These contacts ensured that the police would not hold them (ibid.). He added: 'A lot of the authorities around here are not only not preventing the crime, but they are also part of it' (ibid.). And a high-level Bangladeshi official acknowledged that they are 'well aware that both the authorities in our own country and in Myanmar are involved in the trade [of yaba]' (ibid.). Indeed, in 2015, 26 police officers from Cox's Bazar were on a list the Bangladeshi intelligence agency compiled of people involved in the drug smuggling business (Momin, 2015). As the case of gun smuggling illustrated above shows, it is not just local authorities but in some cases also high government officials that support and profit from the smuggling economy.
In addition to this institutional link, state authorities, particularly officials from law enforcement agencies, appear to be making use of criminal opportunities in the smuggling context as they arise. For instance, newspaper reporting shows that police officers who confiscate yaba only declare part of the haul and keep the rest to sell themselves. For instance, in September 2017, police officers from Cox's Bazar seized 730,000 yaba pills, but only declared 8,000 (Daily Star, 2018). While confiscating some yaba creates the impression that the police are doing something against the drug trade, such measures may also be used to put pressure on independent drug dealers to join the state-linked network and pay a weekly fee, thus ensuring their protection.

CONCLUSIONS
The case of the Myanmar-Bangladesh border areas illustrates a number of dimensions that link armed conflict to the smuggling economy. While the smuggling economy is long-standing, the conflict did create new opportunities and expanded existing ones. On a local level, the state-perpetrated violence in Rakhine resulted in large-scale displacement of people, and the sudden swell of people seeking to cross the Naf river from Myanmar into Bangladesh created ad hoc business opportunities for local fishermen, who charged people for helping them. Meanwhile, having a fairly open border also enabled some refugees to make some money from smuggling yaba across the border, helping them to survive in the camps. Going beyond the activities of such 'little fish', the displacement and resulting situation has also created a growing business opportunity for more organised criminal networks. Such networks 'organise' many of the activities that relate to smuggling. They are in charge of the smuggling of yaba from Myanmar to Bangladesh, using the Rohingya to do the dangerous part; they provide false documents for refugees who can afford it, which allows them to leave the camps and, if they want to, to travel to Malaysia; they traffic women and children out of the camps into forced labour and prostitution; and they organise the smuggling of refugees to Malaysia via boat, with the help of local fishermen.
It is not clear to what extent these businesses are operated by one or multiple networks. However, the networks seem to not only be working across international borders, but also bridging nationalities and ethnic groups. For instance, on the Myanmar side, according to the accounts of victims, both Buddhists and Rohingya are part of the people smuggling operations. On the Bangladeshi side of the border, Bengali people seem to dominate the business, while Rohingya are only used at the bottom of the hierarchy, for instance to transport drugs.
On both sides of the border, criminal networks appear to be closely linked to state authorities, including both high-level officials and local officers. The key role that state actors play in the local smuggling economy became particularly apparent in the context of smuggling illicit goods, such as, most prominently, yaba. With little or no risk of detection and prosecution, the smuggling activities can run much more smoothly and more cost-efficiently than any 'independent' smuggler would be able to achieve.
Conversely, non-state armed groups play little role in smuggling at the Bangladesh-Myanmar border. For criminal networks, non-state armed groups like ARSA are not a relevant partner as they do not control the border or any territory, and are not able to block or facilitate their activities. Further, ARSA members are considered to be 'terrorists' in Myanmar who are actively searched for and association with them carries risks. Hence, ARSA's role seems to be limited to being customers of criminal networks, for instance, when wanting to buy arms.
Not controlling territory, ARSA can be categorised as a rebel, which limits its ability to play a significant role in the smuggling economy. However, compared to the situation for a group like BRN in Southern Florian Weigand -9781789905205 Downloaded from Elgar Online at 11/05/2020 12:53:18PM via free access Thailand, the situation is even more difficult for ARSA. Almost the entire ethnic group ARSA claims to represent, the Rohingya, has been displaced from their territory in Myanmar and is at risk when returning. Hence, it is unlikely for ARSA -or any other exclusively Rohingya organisation -to play any influential role in the smuggling economy. Legitimacy considerations are secondary to territorial constraints. It remains unclear whether building local legitimacy is a priority for ARSA. On the one hand, ARSA is responding to local expectations, resisting the Tatmadaw and, in some cases, helping people to flee from Myanmar. On the other hand, many Rohingya believe ARSA triggered the large-scale violence against the Rohingya. The AA is far more influential than ARSA, and it has networks across the border and controls some territory. This gives it more opportunities to get involved in the smuggling economy, including the smuggling of drugs, and is a more attractive business partner for smugglers. Nonetheless, compared to the systemic involvement of state actors in the smuggling economy of this border region its role appears also to be minor.
Violence and armed conflict, such as that at Myanmar's border with China and at the Thailand-Malaysia border, deters smugglers who prefer stable environments and routes that offer predictability and low risks. Meanwhile, the smuggling economy at Myanmar's Bangladesh border has flourished in spite of the violence that occurred in Rakhine state in August 2017. The conflict increased the demand for people smugglers and the displacement created the demand for goods smuggled in from Myanmar to Bangladesh. Perhaps more surprisingly, the smuggling of transit goods, that only pass through the area, such as yaba, continues on the same if not a larger scale.
Three differences, when compared to Southern Thailand and Kachin State, may help to explain these dynamics. First, violence in Southern Thailand and Kachin State is not intended to push a large number of people out of the country and across the border. Hence, the violence does not spark a new demand for people smuggling or the smuggling of daily consumption goods. Meanwhile, the displaced Rohingya wanted to cross to Bangladesh, requiring the help of fishermen, and in Bangladesh they continue to consume goods from Myanmar that need to be smuggled. The existence of vast refugee camps in Bangladesh in turn has created opportunities for human traffickers. The second difference compared to the other two borders is that there is no alternate route for drug smuggling from the drug production in Myanmar's Shan State to the consumers Florian Weigand -9781789905205 Downloaded from Elgar Online at 11/05/2020 12:53:18PM via free access in Bangladesh's Dhaka that offers a more peaceful connection and that has good roads. Shifting the smuggling route would almost certainly require considerable detours or much more tedious smuggling operations in rough terrain. The third difference is that the episode of violence in Rakhine was short and the activities of non-state armed groups such as ARSA do not affect the smuggling economy greatly. Being closely linked to and protected by state authorities, smugglers continue to have a high level of predictability for their business that enables them to continue their operations.