Legal Aid in Bangladesh: A Matter of Timing

Key words: legal aid, pre-trial detention, prison overcrowding, Bangladesh

Abstract

The recently completed Justice Audit has generated a considerable amount of data from across the criminal justice system in Bangladesh. With the help of this data, it is now possible to demonstrate that making legal aid accessible from the earliest stages of criminal proceedings will make a significant contribution towards reducing the perennial overcrowding in the country’s prisons.

1. Introduction

In 2016, Farzana Akter of the University of Dhaka published an article in the Asian Journal of Criminology, outlining the legal obligations on the part of the Bangladeshi state to make legal aid accessible from the earliest stages of criminal proceedings.[1]

At the time, there was very little data available on the extent to which people were able to access legal aid and at what stages they were able to do so. In 2016, the Ministry of Law, Justice and Parliamentary Affairs commissioned a Justice Audit in order to build an evidence base for justice sector reform. The Justice Audit establishes a system for regularly collecting different types of data, from across the criminal justice system, for the purpose of informing policymaking and programming.[i] With the help of this data, it is now possible to demonstrate the potential effect of the reforms recommended by Farzana Akter, not only in making the Bangladeshi criminal justice system more compliant with both international and domestic law, but also in terms of reducing the perennial overcrowding in the country’s prisons.

2. How to Shorten Pre-Trial Detention

One of the main problems that Farzana Akter points to in her article is ‘excessive and lengthy pre-trial detention.’ The Justice Audit data shows that, at the end of 2016, close to 77% of the prison population was made up of prisoners who had not been convicted.[2] There are two ways of reducing this number. The most obvious is to speed up the court process, so that people in pre-trial detention either get convicted or acquitted. While the Justice Audit points to ways in which this can be achieved, this is not the focus of this particular article. This article instead examines the possibilities of making better use of bail and ensuring that people who are suspected of having committed criminal offences are remanded in custody only when the legal requirements for such detention are met. 

There seems to be at least three ways to further increase the number of arrested people who secure bail either before their first court appearance, at their first court appearance, or very shortly thereafter.

2.1      Use of bail in cases involving alleged non-bailable offences

For people accused of so-called ‘bailable offences’ section 496 of the Criminal Procedure Code offers the accused a right to bail, and the police and the courts must grant it.^[3]In respect of people accused of so-called ‘non-bailable offences’ section 497 (1) states that: ‘When any person accused of any non-bailable offence is arrested or detained…he may be released on bail, but he shall not be so released if there appear reasonable grounds for believing that he has been guilty of an offence punishable with death or transportation for life.’ (Emphasis added).

Among the magistrates who took part in the Justice Audit Practitioner Survey, 76% said that the most common reason for denying bail is that the offence is ‘non-bailable’.[4] The corresponding figure among judges was 70% and among prosecutors 60%.[5] This suggests that even when a suspect can apply for bail, if there is an offence listed on the First Information Report (FIR) that is non-bailable, there is a high risk that bail will be denied merely because the offence falls in the category of so-called ‘non-bailable’ offences. Because of the routine practice of refusing bail in cases supposedly involving non-bailable offences ‘police often act under pressure from the community as well as local elites to…add non-bailable sections [to the FIR] with a view to ensuring detention of the alleged offender.’ (Chowdhury, 2017).

This practice looks to be a misapplication of the law, which in fact creates a rebuttable presumption in favour of bail, also in cases of alleged non-bailable offences.  The Law on Bail states that the ‘underlying principle [is]…to grant bail unless the offence which a person is alleged to have committed is one punishable with death or transportation for life.’^[6] It goes onto explain that the ‘word “may” in [s. 497(1) cited above] no doubt shows that it is discretionary [to grant bail]…This does not, however, mean that the Court has an arbitrary power of refusing bail. It only means that a prisoner’s right to be released on bail can be taken away when the prosecution establishes that if the accused is set free on bail he is likely to commit such acts which may cause hindrance to prosecute the case successfully.’^[7] If there is no evidence causing ‘apprehension of his [the accused] absconding or tampering with the evidence…the Court should almost invariably grant bail.’^[8]

The jurisprudence also makes it clear that detention must never be authorised merely because the police or the prosecution request it; denying bail should never be a routine matter. The requirement that the magistrate authorising detention ‘shall record his reason for doing so’ (section 497(3)) has, by the courts, been taken to mean that ‘the record must show that the Magistrate did apply his mind to the question of further detention in the circumstances of the case, taking into consideration the stage of its investigation and the materials so far gathered. A routine order on police prayer will not be a proper order for further detention.’^[9] In the case of Daulat Singh the court observed that ‘to deprive a person of his liberty is a most serious step to be taken and it is hardly too much to say that every step in the process should show extreme deliberation and care.’^[10]

Regrettably, recent research has found that magistrates ‘often refrain from exercising sound judicial discretion…in rejecting discharge applications and refusing bail at the pre-trial stage.’ (Chowdhury, 2017). Furthermore that ‘it has become a routine practice for magistrates to pass a remand order or an order for detention based on the forwarding by the police officer without perusing the entries of the case diary.’ (Chowdhury, 2017).

Yet the law is unequivocal. If someone is suspected of having committed a non-bailable offence that is not punishable by death or transportation for life, bail should be granted unless there are strong and reasonable grounds for refusing bail. The term ‘non-bailable’ is thus misleading, as it erroneously implies that people accused of offences in this category cannot or should not be granted bail. Quite contrary to what magistrates, judges and prosecutors seem to believe, the fact that there is a non-bailable offence on the FIR or the charge sheet is not at all a valid reason to deny bail.

2.2      Use of bail even when there is a prima facie case against the accused

The second most common reason for denying bail is that ‘the evidence against the accused is strong’.[11] Again, unless the alleged offence is one punishable by death or transportation for life, this practice is at odds with the law. It has been noted that an ‘erroneous impression seems to have crept in lately that in dealing with a bail matter the Court is to form an opinion about the merits of the prosecution’s case as a whole.’^[12] This impression was rejected in Sanyasayya v Public Prosecutor where the judge explained that ‘the Court is not called upon to conduct a preliminary trial of the case and consider the probability of the accused’s guilt or innocence…[but]…to inquire whether the giving of bail, as opposed to the arrest of the accused, might lead to a real reason to suppose that he is likely to tamper with the witnesses who would be called against him.’^[13] Once more, the case law is unequivocal. Bail should be granted unless there is proof that doing so may hinder the investigation. If the courts were to apply this simple rule, as opposed to trying to assess the weight of the substantive evidence, in what often becomes a lengthy de facto preliminary trial, this would probably allow people to obtain bail much earlier.

These apparent misconceptions around when bail should be granted in cases involving alleged non-bailable offences has been described as a ‘gulf between theory and practice’. (Chowdhury, 2017).  This gulf can be narrowed with the help of (a) practice directions from the Chief Justice on how to apply the law, accompanied by (b) appropriate training for magistrates, judges, prosecutors and lawyers, and (c) monitoring of compliance by the Supreme Court. Such measures are likely to result in more people with non-bailable offences on their FIRs or charge sheets securing bail earlier.

2.3      Getting lawyers faster – for everyone, not only for a few

Data from the Justice Audit shows that the police arrested 540,982 people in 2016,[14] and that 406,129 people were admitted to prison as pre-trial prisoners.[15] That means that 134,853 people, or close to 25% of the total number of people arrested, were granted bail either by the police or at their first court appearance. The Court User Survey, conducted by the Bangladesh Bureau of Statistics (BBS) in support of the Justice Audit, shows that about 24% of the respondents who had secured bail got it on their first application.[16] Taken together, this demonstrates that it is possible to secure bail very quickly in the Bangladeshi criminal justice system.

It can probably be assumed that most, if not all of the people who were able to secure bail at this early stage were people who could mobilise lawyers at short notice.[ii] The problem, of course, is that not everyone can hire a lawyer within less than 24 hours. This is why Farzana Akter ‘recommends that the government-operated legal aid system of Bangladesh should adopt a mechanism in order to ensure early access to legal aid in police stations and then in custody before being formally charged with a criminal offence as an essential component of a comprehensive system.’ (Akter, 2017). This would allow people who cannot hire a lawyer quickly to nevertheless submit a bail application at their first court appearance.

However, access to legal services following arrest appears limited. The Justice Audit Court User Survey reveals that among a sample of people who had been arrested, 24% had been able to speak to a lawyer while in police custody.[17] About 85% of that same group of respondents also said that their families knew that they were in police custody.[18] This is significant because when someone is arrested, it is typically the family who makes the arrangements for a lawyer. The sooner the family gets to know about the arrest, the sooner they can hire a lawyer to apply for bail on behalf of the person arrested. Out of those whose families got to know about their arrest, 91% learned about it within 24 hours of the arrest.[19] A dilemma presumably arises concerning those who do not have family nearby who can help find a lawyer. An example might be the growing number of workers who migrate into cities leaving their families in rural homes. This is possibly why we see some of the highest levels of non-convicted prisoners in places with large numbers of migrant workers, such as in Gazipur where 88% of prisoners were non-convicted and in Chittagong, where the figure stood at 91%.[20]

These observations notwithstanding, the Justice Audit also shows that 5,290 people did obtain paralegal aid at police stations from NGOs in 2016.[21] This number is small in a national context, but that is because the NGOs were only able to provide these services in 39 of the country’s 64 districts and within those districts only at some police stations. This is crucial, nevertheless, because it demonstrates that civil society is taking the lead in extending the reach of the legal aid system into police stations. It also shows that in some places, the police are willing to allow legal service providers to meet people who have been arrested. If this is already happening in some police stations, there is probably scope for expanding this practice to more police stations in the future.[iii]

Farzana Akter points out that poor people ‘lack adequate financial resources to bail themselves out…’ (Akter, 2017) and while this is true, the data suggests that almost everyone gets bail, eventually. Data from a study of 6,793 criminal cases pending for one year or more shows that in all but 0.8% of those cases, the accused had obtained bail (GIZ, 2017). Official data from prisons corroborate this. As of June 30, 2016, prisons held 54,992 non-convicted prisoners.[iv] Out of these 15,252 (just under 28%) had been in prison for one month or less (Department of Prisons, 2017).  Another 14,604 (close to 27%) had been in prison for one to three months (Department of Prisons, 2017). What this means is that most people who are remanded in custody remain there for less than three months, after which they are released on bail. These figures can also be triangulated with the data from (a) the JA Prison Baseline Data, which shows that as of December 31, 2016, close to 66% of non-convicted prisoners had been in custody for less than six months,[22] and (b) the JA Court User Survey, which shows that of all respondents who had served prison time in connection with their case, over 82% had been released on bail within six months.[23]

What we can draw from this is that the importance of financial resources is perhaps not so much in relation to whether bail is granted or not, but rather when it is granted.[v]

2.3.1   Problem: legal aid is not proactive

The state legal aid system today is reactive. When clients are referred to it by NGOs, local authorities, prisons, or when people approach the legal aid offices, the system will respond, and it does this well. Over 95% of all applications submitted to legal aid offices across the country are approved.[24] However, as Farzana Akter points out, by being only reactive, the state legal aid system misses the opportunity to provide critical assistance at the early stages of criminal proceedings (Akter, 2016). The JA Legal Services Baseline Data illustrates the point. The largest portion of legal aid applications come from prisons. This means that the prisoner in question has been unrepresented long enough for the prison authorities to take notice, add the prisoner to a list of unrepresented prisoners, and notify the district legal aid officer.[25] In such a scenario, by the time the District Legal Aid Office assigns an advocate, the applicant’s remand warrant will have been renewed several times, and several opportunities to apply for bail will already have been missed.

2.3.2   Problem: legal aid is slow

In the Court User Survey, about 20% of the respondents who had a lawyer believed that they had the ability to find a legal aid lawyer,[26] yet less than 9% had actually tried to find a legal aid lawyer.[27] When asked why they preferred a paid lawyer, 34% said it was because legal aid lawyers ‘work very slowly’.[28] Only 16% of respondents believed that a legal aid lawyer would be as good as a privately hired lawyer.[29]

This should not necessarily be taken as criticism of the professionalism of the legal aid lawyers themselves. It may be an effect of what might be called ‘the monetisation’ of the criminal justice system. At every step of the criminal process, lawyers are able to offer court staff and other officials unofficial fees (or ‘speed’ money) in exchange for getting their cases or applications prioritised so as to speed up the process for their clients.[vi] With a paying client, this presents no problem as such expediency fees are factored into the billing of the client. With a legal aid client, however, paying unofficial fees becomes difficult, maybe impossible as the client’s family has little or no money to cover such costs (if they did, they would have hired a private lawyer in the first place). If a lawyer is unwilling to pay the fees, he or she may eventually get the documents anyway, but it will almost certainly take longer than if a fee had been paid.[vii]  Chowdhury notes how ‘panel lawyers…are unable to pay speed money to the court-staff as a result of which they are disinclined to assist with the disposal of cases of poor litigants.’ (Chowdhury, 2017). The inability to pay the expediency fees may be a stronger factor behind the perceived delays in legal aid cases than the lack of commitment by the lawyers.

The Baseline Data confirms the findings of the Court User Survey. In order to qualify for legal aid, an applicant must meet the criteria of the means test in Legal Aid Services Act.[viii] Anyone who falls below the official poverty line would be a person who qualifies for legal aid. The poverty rate in 2016 was 24.3% (World Bank, 2016). If it is assumed that the poverty rate among people who are arrested is at least as high as it is among the general population, then at least 24.3% of people arrested in 2016 qualified for legal aid. Based on these very conservative estimates, that would mean that at least 131,459 arrested persons qualified for legal aid and presumably needed it because they were poor.[30]  The data, however, showed that nationwide, only 22,495 people applied for legal aid in criminal matters in 2016, i.e. about 17% of those who qualified.

The data indicates that many people who qualify for legal aid, who apparently need it, and who know how to access it, still chose not to use it. For a person who is arrested, getting legal help quickly is important anywhere in the world, but for a person detained in a Bangladeshi prison, it is perhaps more urgent than in most other jurisdictions (Tamim, 2016).[ix] For a person in that situation, every hour and every minute counts. As long as there is reason to believe that a private lawyer can get a loved one out sooner than a legal aid lawyer can, poor people will continue to sell their belongings and borrow as much money as they possibly can to hire private lawyers.

2.3.3   Solution: a duty advocate scheme

In theory, there is nothing that bars an accused person from representing himself and applying for bail. In practice, however, that happens only very rarely, if at all. Bail is almost only ever granted if an advocate files the application for bail on behalf of a client.[x] Unless this practice is changed, getting bail for more people at or before their first appearance will require that advocates are available to apply for bail also for all those who are not able to hire an advocate within 24 hours.

A duty advocate scheme will ensure that everyone is represented at their first court appearance. Given the shortage of available legal aid lawyers,[31] this duty advocate scheme will have to be supported by teams of paralegals. Paralegals can prepare bail applications based on a standardised form, ideally already at police stations (the JA shows that this is possible) or at court, as soon as the detainees are delivered in the morning at the court holding cells. With the bail applications already prepared, the duty advocate can file them in court for each of the unrepresented persons due to appear that day. This will ensure that everyone gets a chance to apply for bail at his or her first appearance. The same scheme can, of course, also serve persons already in prison, coming for subsequent court appearances. If they are unrepresented, and if they wish to file a bail application, the duty advocate scheme will allow them to do so.[xi]

As noted earlier, a significant cause of delay is that legal aid lawyers are unable to access the documents they need to adequately represent their clients without paying unofficial expediency fees. The Case Coordination Committees, which exist in almost half the districts, should identify pathways that allow legal aid duty advocates and the paralegals to immediately access all documents necessary to prepare bail applications without paying unofficial fees.

4. The Effect on Prison Overcrowding

Farzana Akter explains why Bangladesh is obliged, under both international and domestic law, to ‘provide early access to legal aid in criminal proceedings.’ Based on that she recommends ‘that the government adopt a mechanism in order to ensure early access to legal aid in police stations and then in custody prior to being formally charged…’ (Akter, 2016). The Justice Audit data reveals that, aside from the legal reasons cited by Farzana Akter, there is also a practical reason for adopting such a mechanism: it will substantially reduce the number of pre-trial prisoners, thereby easing the current levels of severe congestion in Bangladeshi prisons.[xii]

The Justice Audit shows that the bulk of people in prison are waiting for their trial, and almost all of them will eventually end up getting bail. One way of reducing the overcrowding, therefore, is to ensure that pre-trial prisoners get bail sooner. A person who gets bail at their first appearance never has to spend a single day in prison and will never take up any space in prison.

The question then arises, out of today’s standing pre-trial prison population, how many could realistically obtain bail at their first appearance? As noted earlier, recent data from the Department of Prisons reveal that well over half of pre-trial prisoners get bail within 90 days, and out those, about half get bail within 30 days (Department of Prisons, 2017). That means that they get bail on their second or third appearance.[xiii] Quite clearly, in these cases, there are no major legal objections to bail being granted – if there were, it would not happen so relatively quickly.

Furthermore, research has found that it takes, on average, approximately 100 days to complete a police investigation and submit the report to the court (GIZ, 2017). This means that in respect of most of the people who are granted bail within 90 days, the decision to grant bail is taken based on the same information that was available to the magistrate at the first appearance, i.e. the FIR and the police report.[32] It can be deduced from this that these are mostly cases in which it took a couple of weeks for the family of the accused to hire a lawyer to apply for bail, and in the meantime, the accused spent time taking up space in prison. If bail had been applied for at first appearance, there is a high likelihood that it would have been secured at that stage since the magistrate, in most cases, will not have access to more information about the cases on the third appearance than on the first.

So, while it is impossible to say, with certainty, by how much faster legal aid might reduce prison overcrowding, it can safely be said that at least half of all pre-trial prisoners could probably have obtained bail at their first appearance if (a) people arrested had received immediate legal aid and if (b) courts applied the presumption in favour of bail, as set out in the applicable caselaw.

What would that mean in practice? If we use the data from the end of 2016 for purposes of illustration, it would mean that there would only have been 29,312 pre-trial prisoners in the standing population. Together with convicted prisoners, instead of having a total standing prison population of 76,614 there would have been 47,188 on December 31, 2016. This would still represent 128% of capacity, but considerably less than the over 208% of capacity that the actual number represented. The drop itself, from 76,614 to 47,188 would be a drop of over 38%.

This is what would happen in Farzana Akter’s recommendation to ensure early access to legal aid in police stations and then in custody prior to being formally charged’ (Akter, 2016) were to be implemented.  

This in turn would have consequences for both the prison administration and for prisoners. The Justice Audit calculates some of the financial savings that a reduction in the number of prisoners would give rise to within the prison administration.[xiv] More importantly, the reduced levels of overcrowding would result in improved humanitarian conditions in prisons.[xv]

Compliance with Ethical Standards

Funding: The Justice Audit in Bangladesh is funded by the Governments of Bangladesh, Germany and the United Kingdom

Ethical approval: This article does not contain any studies with human participants or animals performed by any of the authors.

Informed consent: Not applicable.

References

Akter, Towards a Comprehensive Legal Aid System in Bangladesh: The Need for Early Access to Legal Aid in Criminal Proceedings, Asian Journal of Criminology (2016) 11:65–82

Bangladesh Justice Audit (2016) at https://bangladesh.justiceaudit.org (accessed September 5, 2018)

Chowdhury, Consultancy Services for Undertrial (UT) Prisoners in Bangladesh, Final Report, June 28, 2017, available at https://bangladesh.justiceaudit.org/wp…/Final-Report_UT-Study_30-Jan-2018-1.docx (accessed September 5, 2018)

Department of Prisons, Prison Population Statistics (2017)

Law on Bail, Dhaka Law Reports, Fourth Edition (2016)

Summary Report on Situation Analysis of case backlog in 5 Districts (unpublished) GIZ (2017)

Mashfiq Tamim, Prisoners’ Right in Bangladesh: Laws, Reality and Solutions, Bangladesh Law Journal, Vol. 5, Jul – Dec, 2016 available online at http://www.lawjournalbd.com/2016/08/prisoners-right-in-bangladesh-laws-reality-and-solutions/ (accessed September 5, 2018)

Human rights implications of overincarceration and overcrowding, Report of the United Nations High Commissioner for Human Rights, Human Rights Council, Thirtieth session, available at http://undocs.org/A/HRC/30/19 (accessed September 5, 2018)

---

[1] Towards a Comprehensive Legal Aid System in Bangladesh: The Need for Early Access to Legal Aid in Criminal Proceedings, Asian Journal of Criminology (2016) 11:65–82

[2] Prison Baseline Data, Justice Audit (2016). While some of the 58,624 non-convicted prisoners may have seen their trial begin and therefore technically be ‘under-trial’ as opposed to ‘pre-trial’, the data suggest that the vast majority of the non-convicted prisoners were in fact pre-trial. See page X of this article.

^[3] ‘When any person other than a person accused of a non-bailable offence is arrested or detained…such person shall be released on bail.’

[4] Question 15, Cross Tabulation of the Practitioner Survey, GJG (2018)

[5] Ibid.

^[6] Law on Bail, Dhaka Law Reports, Fourth Edition (2016). pp 32 – 33.

^[7] Law on Bail, op cit. 34 – 35.

^[8] Ibid.

^[9] Khairathi Ram 23 CrLJ 913 (914). See also Kundan Lal 32 CrLJ 785 (788); Hidayat Begum 52 CrLJ 233; Bir Bahadur Pratap Singh v DM Azamgarh AIR 1959 (3) 384, cited in Law on Bail op. cit. p. 21

^[10] In the matter of the petition of Daulat Singh, ILR 14 All. 45 at 47 cited in Law on Bail, op. cit. p 34

[11] Question 15 Practitioner Survey, Justice Audit (2016)

^[12] Law on Bail, op. cit. 83

^[13] Madras High Court (19 December 1924) 26 CrLJ 1593, cited in Law on Bail, op. cit. p 37

[14] Police Baseline Data, Justice Audit (2016)

[15] Prison Baseline Data, Justice Audit (2016)

[16] Question 60, Court User Survey, BBS (2018)

[17] Question 48, Court User Survey, BBS (2018)

[18] Question 50, Court User Survey, BBS (2018)

[19] Question 51, Court User Survey, BBS (2018)

[20] Prison Baseline Data, Justice Audit (2016)

[21] Legal Services Baseline Data, Justice Audit (2016)

[22] Prison Baseline Data, Justice Audit (2016)

[23] Question 53, Court User Survey (2017)

[24] Legal Services Baseline Data, Justice Audit (2016)

[25] Out of the 22,495 applications for legal aid in criminal matters that reached District Legal Aid Offices in 2016, 9,694 originated from applicants in prison. Legal Services Baseline Data, Justice Audit (2016).

[26] Question 25, Court User Survey, BBS (2018)

[27] Question 27, Court User Survey, BBS (2018)

[28] Question 28, Court User Survey, BBS (2018)

[29] Question 29, Court User Survey, BBS (2018)

[30] Police Baseline Data, Justice Audit (2016)

[31] See Legal Services Baseline Data, Justice Audit (2016).

[32] Known in Bangladesh as the ‘police forwarding’.

---

[i] The Justice Audit collects data on the whole criminal justice system viewed as an interconnected set of processes and provides a diagnosis based on the data. Data are gathered from a range of sources and triangulated – hard (administrative) data with citizen survey and practitioner and system user interview data. The data are organised in a quantitative baseline and then visualised to provide an accessible display of quantitative and qualitative data as an open source document. See generally http://www.governancejustice.org/justice-audit and for the Bangladesh Justice Audit, see https://bangladesh.justiceaudit.org

[ii] It is very rare for magistrates to consider bail unless an application for bail is made, and in the Bangladeshi criminal justice system, it is practically difficult for an unrepresented person to apply for bail.

[iii] Chowdhury notes, however, that the ‘frequent transfer of police officers from sadar police stations with whom the paralegals had forged a good working relationship is a major impediment in providing legal assistance…at an early stage.’ Chowdhury op. cit. p. 37

[iv] In Bangladesh usually referred to as ‘under-trial’ prisoners, even though in respect of the vast majority of them, charges have not been framed and the trial has not commenced, so they are in reality ‘pre-trial’ prisoners. 

[v] In fact, a good lawyer in Bangladesh would even be able to secure bail before an anticipated arrest. However, that will not come cheaply and is an option only for the very wealthiest. See section 498 of the Criminal Procedure Code.

[vi] A typical example is when a lawyer accepts a case and must obtain certified copies of the necessary case documents from a court (known as ‘folio’). The official process for this is so cumbersome and time consuming that bypassing it, often with the help of an informal expediency fee, is routine.

[vii] In the JA Practitioner Survey, 27% of lawyers and paralegals cited ‘corruption’ as one of the things that frustrate them the most in their work. This was higher than for any other practitioner category. Question 26, Practitioner Survey, Justice Audit (2017).

[viii] The current annual income limit for eligibility is 100,000 Taka. However, given the difficulties in assessing people’s income in an economy with an informal sector as large as that in Bangladesh, discretion has to be exercised in applying the test. There is also a range of alternative criteria that can be cited to prove eligibility. In practice, it is a non-issue as all applications are approved anyway, with only the tiniest exception. Legal Services Baseline Data, Justice Audit 2016

[ix] For a description of prison conditions in Bangladesh today, see, e.g. Mashfiq Tamim, Prisoners’ Right in Bangladesh: Laws, Reality and Solutions, Bangladesh Law Journal, Vol. 5, Jul – Dec, 2016 available online at http://www.lawjournalbd.com/2016/08/prisoners-right-in-bangladesh-laws-reality-and-solutions/

[x] An arrested person who is not represented is unlikely even to get to make a physical appearance before the magistrate. Instead his custody warrant will often be signed, as a matter of routine, in chambers. See Chowdhury op. cit. pp. 34 and 52

[xi] This, of course, is not something new. Duty solicitor schemes are operating in many countries all over the world precisely with the aim of ensuring that everyone who needs a lawyer can get one in time for their first court appearance.

[xii] At the end of 2016, prisons in Bangladeshi had the combined capacity to hold 36,742 people. The standing prison population was 76,614. Out of these, 58,624 were awaiting trial. Prison Baseline Data, Justice Audit (2016).

[xiii] The maximum duration for which a person can be remanded in custody is 15 days (Section 167 of the Criminal Procedure Code). This means that even if someone was unable to apply for bail at first appearance, new opportunities to do so arise at least every 15 days. A bail application can also be filed at any time between the scheduled appearances. The challenge for the legal aid system is to identify people in need of an advocate so that bail applications can be filed earlier.

[xiv] The current cost of feeding a prisoner is estimated to be 10 Bangladeshi taka (approximately 12 US cents) per day. See https://bangladesh.justiceaudit.org/national-data/key-measures/prisons/

[xv] Human rights implications of overincarceration and overcrowding, Report of the United Nations High Commissioner for Human Rights, Human Rights Council, Thirtieth session, available at http://undocs.org/A/HRC/30/19