Witness protection and ending impunity; making military justice systems human rights compatible

The investigative phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Prosecutorial arrangements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Conduct of trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. The design of a formal witness protection programme . . . . . . . . . . . . . . . . . . . . . . . . . . . .

The structure of a witness protection programme . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Protection measures short of relocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Relocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Innovative approaches to relocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Making military justice systems human rights compatible . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Legal framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Case studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V. Conclusions and recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

The present report focuses on outstanding country visit requests and provides a brief note on visits undertaken since the previous report to the General Assembly. The bulk of the report addresses two neglected but vital dimensions of the struggle to combat impunity for extrajudicial executions: the provision of effective witness protection arrangements; and the importance of ensuring that military justice systems are compatible with human rights standards. In many States, witness protection programmes are wholly inadequate and military justice systems are structured in ways that promote impunity for killings. The report surveys best practices in both areas.

While the challenges of every country are unique, there is much to learn from how others have confronted similar challenges. This report draws on both the own experiences of the Special Rapporteur and on recent research to survey some approaches to providing witness protection and reforming military justice systems.

In the preparation of the report, the Special Rapporteur is grateful to Mr. William Abresch of the Project on Extrajudicial Executions at New York University School of Law, who has provided superb assistance and advice, as well as to Mr. Euan MacDonald for an excellent review of issues relating to witness protection.

Since the previous report of the Special Rapporteur to the General Assembly, the Governments of Afghanistan and the United States of America have issued invitations at my request. As of August 2008, I have requested visits to 34 countries and the Occupied Palestinian Territories. Only 11 of those — Afghanistan, the Central African Republic, Guatemala, Israel,¹ Kenya,² Lebanon, Nigeria, Peru, the Philippines, Sri Lanka and the United States of America — have actually proceeded with plans for a visit. The visit to Peru was cancelled, and the Palestinian Authority issued an invitation.

1. Israel agreed to issue an invitation to visit in connection with the visit of the Special Rapporteur to Lebanon after the conflict in 2006. To date, however, it has yet to respond favourably to repeated requests to visit in relation to issues relating to the Occupied Palestinian Territories. ↩
2. The Government has issued an official invitation and the mission is proposed to take place in December 2008. ↩

The responses of the remaining 23 countries have ranged from complete silence through formal acknowledgement to acceptance in principle but without meaningful follow-up, to outright rejection. In some cases, the relevant requests were first made some eight years ago.

States which have so far failed to respond affirmatively to requests for a visit are: Algeria, Bangladesh, China, El Salvador, Guinea, India, Indonesia, Iran (Islamic Republic of), Israel, Lao People’s Democratic Republic, Nepal, Pakistan, Russian Federation, Saudi Arabia, Singapore, Thailand, Trinidad and Tobago, Togo, Uganda, Uzbekistan, Venezuela (Bolivarian Republic of), Viet Nam and Yemen.

Seven members of the Human Rights Council have failed to issue requested invitations, most of which have been pending for several years: Bangladesh, China, India, Indonesia, Pakistan, Russian Federation and Saudi Arabia. This is noteworthy in view of the pledges offered in connection with elections to the Human Rights Council and of provisions in the code of conduct for special procedures mandate-holders of the Human Rights Council urging all States to cooperate with, and assist, the special procedures in the performance of their tasks.³

3. See A/HRC/5/1/Add.1. ↩

The Special Rapporteur visited Brazil, including Brasília, São Paulo, Rio de Janeiro, and Pernambuco, from 4 to 14 November 2007. The major problem identified was the very high rate of homicide, accompanied by relative impunity. Between 45,000 and 50,000 homicides are committed every year in Brazil. Although these killings have sown widespread fear and insecurity among the general population, remarkably little is done in the vast majority of such cases to investigate, prosecute and convict the culprits. Other major concerns include (a) killings by vigilante groups, death squads, extermination groups and militias; (b) prison killings; and (c) killings by police.

The Special Rapporteur visited the Central African Republic, including Bangui, Bossangoa, and Paoua, from 31 January to 7 February 2008. The preliminary report of the Special Rapporteur identified killings by the military, combined with the de facto impunity of those responsible, as the main issue. Other problems included killings by soldiers for personal reasons or in connection with corruption efforts, killings in police custody or detention facilities, and killings of those accused of witchcraft (“sorcery” or “charlatanism”).

The Special Rapporteur visited Afghanistan from 4 to 15 May 2008, including Helmand, Kabul, Kandahar, Kunar, Nangarhar, Jowzjan and Parwan. Afghanistan is engulfed in an armed conflict in which large numbers of avoidable killings of civilians are being tolerated. If the situation is to improve police killings must cease; widespread impunity within the legal system for killings must be rejected; the killing of women and girls must end; and the international military forces must ensure real accountability for their actions. Responding to the latter concerns, a spokesman of the North Atlantic Treaty Organization attacked “the substance and the overall tone” of the statement by the Special Rapporteur, calling it “inaccurate and unsubstantiated”. The spokesman did not, however, consider it necessary to address any of the substantive concerns the Special Rapporteur identified. This is regrettable and the final report of the Special Rapporteur will deal with that issue in detail.

The Special Rapporteur visited the United States of America from 16 to 30 June 2008, including Washington, New York, Montgomery (Alabama), and Austin (Texas). Without significant reforms to the criminal justice system innocent people are clearly at risk of execution. Measures are needed to enhance judicial independence, ensure adequate defence counsel, and review capital cases on the merits at the appellate level. Other recommendations concern improving the transparency of the military justice system, ensuring accountability for private security contractors, and building on existing efforts to provide reparations to victims in armed conflict. The Special Rapporteur also addressed deaths in immigration detention facilities, due process concerns in death penalty cases against “alien unlawful enemy combatants”, and the deaths of detainees at the U.S. military facility at Guantánamo Bay.

The successful prosecution of those responsible for extrajudicial executions is difficult, if not impossible, in the absence of effective witness protection programmes. All too rarely are prosecutions built on painstaking forensic and other investigative work which would reduce the need to rely upon witnesses. If witnesses can be easily intimidated, if they and their families remain vulnerable, or if they sense that the protections offered to them cannot be relied upon, they are unlikely to testify. As a result, it is often the case that the only people willing to take the risk of testifying are the victims’ family members. Usually, however, they are poorly placed to provide the most compelling evidence against the perpetrators. Ending impunity for killings thus requires institutionalizing measures to reduce the risks faced by witnesses who testify. Yet it is often the States that have the biggest problems that also have the least adequate witness protection arrangements.⁴ Similarly, many States have sophisticated programmes to protect witnesses to murders involving organized crime, but they devote much less attention to protecting witnesses to murders implicating their military or police forces. This latter challenge requires distinctive solutions, in part because relying on the police to provide protection may itself compromise at least the appearance and often the reality of protection.

4. See A/HRC/8/3/Add.4, paras. 19 and 21 (d) (Brazil); A/HRC/4/20/Add.2, paras. 51 and 63 (Guatemala); A/HRC/8/3/Add.2, paras. 52-54 and 71 (Philippines); and E/CN.4/2006/53/Add.5, para. 56 (Sri Lanka). ↩

The central importance of effective witness protection programmes in efforts to combat extrajudicial executions has been generally overlooked by the international community and there have been all too few efforts to encourage States to devote the necessary efforts and resources to the issue. This report thus aims to highlight examples of global best practice and identify some of the key issues that need to be addressed in the design of effective programmes. The resulting survey is far from comprehensive and relies heavily upon scholarly works, reports commissioned by governments, and observations from country visits by the Special Rapporteur.⁵

5. Particularly helpful resources included: Law Commission of India, Consultation Paper on Witness Identity Protection and Witness Protection Programmes (Aug. 2004); K. Dedel, Witness Intimidation (US Department of Justice, July 2006); N. Fyfe and J. Sheptycki, “International Trends in the Facilitation of Witness Co-operation in Organized Crime Cases”, 3 European Journal of Criminology (2006) 319; Council of Europe, Report on Witness Protection (Best Practice Survey) (1999); and P. Finn and K. M. Healey, Preventing Gang- and Drug-Related Witness Intimidation (National Institute of Justice, Nov. 1996). ↩

The starting point for effective programmes is to acknowledge that the successful prosecution of killers is in the best interests of the society. Witness protection should thus not be seen as a favour to the witnesses who are in fact often making immense personal sacrifices on behalf of society. The provision of adequate assistance to witnesses, family members, and others against whom retaliation is feared, is thus a necessary condition for breaking the cycle of impunity. Such assistance must be provided in a constructive and pragmatic spirit. Dogmatic approaches must be avoided. For example, witnesses may consider the whole of the security forces to be systematically engaged in abusing rights, while the government may believe that the security forces are basically reliable despite the presence of rogue officers in some specific units. In such cases, it is tempting for the government to adopt a dogmatic attitude and insist that any protection be provided by the security forces, thus refuting assumptions that systemic problems exist. A pragmatic approach, however, would place a premium on winning the trust of the witnesses and would provide protection in the most effective and acceptable manner.

Witness protection cannot be viewed as an isolated challenge. Rather, it must be seen as a crucial part of a comprehensive system designed to effectively investigate, prosecute, and try perpetrators of human rights abuse. Witness protection will be ineffective if the other components of the criminal justice system are not also functioning well. Moreover, a holistic approach will help to identify ways in which less reliance can be placed on testimonial evidence and methods of witness protection which do not require full-blown witness protection programmes. Every step of the process from investigation through conviction and incarceration should be analysed to identify ways in which witnesses are placed at risk and potential reforms designed to limit those risks.

Already at this phase there can be problematic disclosures of witness identities. When this risk is foreseen by witnesses, they may simply choose not to speak with investigators. Conversely, safeguarding the identity of a witness at this early point enhances the potential for safely obtaining testimony at trial without resorting to a full-blown witness protection programme. A British report identified several policing methods designed to limit risks in the early phases of the investigation:
• Police should give only minimal information about witnesses over their radios
• Police should not visit witnesses on the day of the incident, and should either encourage the witness to come to the station to give a statement, send a plain-clothes officer to the home of the witness, or conduct a number of house-to-house calls in order to prevent the witness from being singled out. The choice should be left to the witness
• Physical screens hiding the witness from the suspect should be used in all identity parades
• A suspect should not be released when the witness is in the vicinity of the police station
• Police should warn witnesses of the risks of potential retaliation in a pragmatic manner that facilitates both their security and their cooperation
• A contact officer should be provided, so that any intimidation can be reported and acted upon immediately
• In some cases, police can hold suspects on remand and restrict their telephone rights to prevent them from contacting witnesses or accomplices to encourage intimidation⁶

6. W. Maynard, Witness Intimidation: Strategies for Prevention, Crime Detection and Prevention Series No. 55, London: Home Office Research Development and Statistics Directorate (1994), pp. 26-30. ↩

In the special case of crimes implicating police or other state agents, the institutional affiliation of investigators may lead to an actual or perceived risk that witness identities will be improperly disclosed. This risk may be mitigated through respect for international norms requiring the removal of suspected perpetrators from positions of power or control over witnesses.⁷ The existence of respected police internal affairs units might also foster confidence on the part of witnesses. But breaking the cycle of impunity will often require an independent investigative unit devoted to solving crimes involving members of the police or security forces.⁸ Thus, for example, the Department of Special Investigation in Thailand was established partly to facilitate the investigation of such cases. Another approach is to remove responsibility for this kind of investigation from the normal investigative police to some other existing body. In Brazil, the Public Prosecutor’s Office sometimes directly investigates murders implicating the police, even though criminal investigations are normally conducted by police detectives.⁹

7. See Principles on the Effective Prevention and Investigation of Extra-legal, Arbitrary and Summary Executions, para. 15. For the test of the Principles, see Official Records of the Economic and Social Council, 1989, Supplement No. 1 (E/1989/89), res. 1989/65, annex. ↩
8. See para. 22 of the Basic Principles on the Use of Force and Firearms by Law Enforcement Officials, adopted at the Eighth United Nations Congress on the Prevention of Crime and the Treatment of Offenders, held in Havana, 27 August to 7 September 1990 (A/CONF.144/28/Rev.1); and para. 11 of the Principles on the Effective Prevention and Investigation of Extra-legal, Arbitrary and Summary Executions, para. 11 (see note 7 above). ↩
9. See A/HRC/8/3/Add.4, para. 21 (f). ↩

Reforms designed to improve the collection of non-testimonial evidence should also be considered. In his visits to countries, the Special Rapporteur observed that investigators often lack the training and resources required to gather forensic evidence — e.g. to match a bullet to a particular gun — leading to excessive reliance on witness testimony.¹⁰ Insofar as the capacity to use forensic evidence can be increased, both the importance of witness testimony and the risks posed to witnesses may be reduced.

10. See A/HRC/8/3/Add.4, para. 21 (c) (Brazil); A/HRC/4/20/Add.2, para. 48 (Guatemala); E/CN.4/2006/53/Add.4, paras. 89 and 105 (b) (Nigeria); A/HRC/8/3/Add.2, para. 55 (Philippines); and E/CN.4/2006/53/Add.5, para. 56 (Sri Lanka). ↩

The role played by prosecutors in the criminal justice system may also lead to real or perceived risks that prevent the safe cooperation of witnesses. Prosecutors generally have close professional relationships with the police, who usually gather the evidence on which their cases are based and often testify in court. These close relationships can lead witnesses to perceive, sometimes correctly, that prosecutors will not push the case aggressively (which makes becoming a witness a pointless risk) or will even be likely to improperly disclose witness identities or locations to implicated members of the police force. There are a number of possible responses to this problem. One is to assign a few prosecutors to work solely on cases involving the police or other government agents in human rights abuses, thus seeking to minimize the normal professional solidarity between prosecutors and police. In one Brazilian city I observed that a prosecutor who played this role had earned the trust of victims of police violence despite their wariness of the prosecutor’s office as a whole.¹¹ Another approach is to establish a separate institution for the prosecution of police. In the Philippines, for example, a special institution was established to investigate and prosecute crimes and other misconduct committed by public officials and its independence from the executive branch was constitutionally guaranteed.¹²

11. For preliminary observations on Brazil, see A/HRC/8/3/Add.4. ↩
12. See Constitution of the Republic of the Philippines (1987), art. XI, sect. 13; see also A/HRC/8/3/Add.2, paras. 56-58. ↩

The manner in which trials are conducted has both indirect and direct implications for witness protection. Scheduling and venue decisions, for example, can have a major bearing on witness participation and protection, but their importance is often overlooked. Opportunities for witness intimidation increase when trials are repeatedly delayed or are not held on consecutive days. Even if the overall workload of the courts makes delays inevitable in many cases, it is worth considering whether the kinds of cases that generally place witnesses at risk can be expedited.¹³ Similarly, judges should avoid summoning witnesses on days other than when they have been scheduled to testify.¹⁴ The way in which rules on venue changes are interpreted is also important. Whether or not they are in a formal witness protection programme, witnesses often relocate to avoid retaliation; if the trial venue can be changed to accommodate their need to remain at a distance from where the perpetrator or his associates live, this can facilitate witness participation.

13. For an example of one approach, see Philippines Administrative Order No. 25-2007, “Designation of Special Courts to Hear, Try and Decide Cases Involving Killings of Political Activists and Members of Media” (1 March 2007); see also A/HRC/8/3/Add.2, para. 59. ↩
14. See also the 154th Report of the Law Commission of India, On the Code of Criminal Procedure 1973 (1976), pp. 43-44. ↩

Permitting witnesses to give testimony anonymously is one of the most controversial but important ways of protecting them. It avoids the need for relocation and other protection measures and can be arranged in such a way as to offer relatively assured protection. But anonymous testimony risks violating the defendant’s right to fair hearing and “to examine, or have examined, the witness against him”.¹⁵ For that reason, courts have subjected such schemes to close scrutiny,¹⁶ while at the same time leaving some space for approaches designed to protect both the due process rights of defendants and the lives of witnesses. Procedures and criteria for permitting anonymous testimony that attempt to respect these rights have been developed by a large number of states,¹⁷ as well as by the International Criminal Tribunal for the Former Yugoslavia (ICTY).¹⁸ In general, the following criteria applied in New Zealand appears reasonably representative of efforts to achieve an appropriate balance:
(4) The Judge may make a witness anonymity order if satisfied that —
(a) The safety of the witness or of any other person is likely to be endangered, or there is likely to be serious damage to property, if the witness’s identity is disclosed; and
(b) Either —
(i) There is no reason to believe that the witness has a motive or tendency to be untruthful, having regard (where applicable) to the witness’s previous convictions or the witness’s relationship with the accused or any associates of the accused; or
(ii) The witness’s credibility can be tested properly without disclosure of the witness’s identity; and
(c) The making of the order would not deprive the accused of a fair trial.
(5) Without limiting subsection (4), in considering the application, the Judge must have regard to —
(a) The general right of an accused to know the identity of witnesses; and
(b) The principle that witness anonymity orders are justified only in exceptional circumstances; and
(c) The gravity of the offence; and
(d) The importance of the witness’s evidence to the case of the party who wishes to call the witness; and
(e) Whether it is practical for the witness to be protected by any means other than an anonymity order; and
(f) Whether there is other evidence which corroborates the witness’s evidence.¹⁹

15. See International Covenant on Civil and Political Rights, General Assembly resolution 2200 (XXI), annex, art. 14. ↩
16. See, esp., the judgements of the European Court of Human Rights in Kostovski v. The Netherlands (1990) 12 EHRR 434; Windisch v. Austria (1991) 13 EHRR 281; Isgro v. Italy (1991) Yearbook of the ECHR, p. 151; Doorson v. The Netherlands, Reports of Judgments and Decisions 1996-II, p. 446; Van Mechelen v. The Netherlands, Reports of Judgments and Decisions 1997-III, p. 691; Fisser v. The Netherlands 13 Human Rights Case Digest (2002) pp. 65-68. ↩
17. Two illustrative examples are Portugal and the United Kingdom. In Portugal, Section 16 of Act. No. 93/99 of 14 July 1999 lays down the conditions precedent for any grant of anonymity to a witness: the testimony must relate to a specific group of crimes; the witness, his relatives or other persons in close contact with him must face a serious danger of attempt against their lives, physical integrity, freedom or property of a considerably high value; the credibility of the witness is beyond reasonable doubt; and the testimony provides a relevant contribution to the evidence in the proceedings. The situation in the United Kingdom is more complicated. In 1995 the Court of Appeal upheld full anonymity, provided that certain criteria were satisfied. R v. Taylor (1995), Crim. LR 253. In June 2008, however, the House of Lords held that the use of anonymous witnesses prevented the accused from adequately examining his accusers and thus violated the right to a fair trial. R v. Davis [2008] UKHL 36. The Government immediately adopted new legislation authorizing the continued use of anonymous witnesses in certain circumstances. See Criminal Evidence (Witness Anonymity) Act 2008. The human rights implications of the legislation have been strongly criticized. ↩
18. The ICTY held that five criteria must be satisfied to permit the use of protective measures: 1) A real fear for the safety of the witness or his or her family; 2) The testimony must be important to the prosecutor’s case; 3) The Court must be satisfied that there is no prima facie evidence that the witness is untrustworthy; 4) The ineffectiveness or non-existence of a witness protection programme; 5) If a less restrictive measure would give the desired level of protection, it must be used. See Tadić (IT-94-1), Decision on the Prosecutor’s Motion Requesting Protective Measures for Victims And Witnesses (10 August 1995), paras. 53-86. ↩
19. See section 13C of the amended New Zealand Evidence Act of 1908 (amended 1997) available at www.legislation.govt.nz/act/public/1908/0056. ↩

But witness anonymity throughout a trial requires more than an appropriate legal framework. The trial itself is one of the most dangerous phases in the criminal justice process for witnesses, when they are often out in the open and thus more susceptible to intimidation; and, where a decision has been made to grant either full anonymity or confidentiality, the possibilities of their identities being disclosed are high during their visits to the courtroom, whether or not they are actually on the witness stand.²⁰

20. See P. Finn and K. M. Healey, Preventing Gang- and Drug-Related Witness Intimidation (U.S. Department of Justice, National Institute of Justice, Nov. 1996), p. 13. ↩

One approach designed to avoid these dangers is to permit the use of anonymous hearsay testimony in criminal trials. But this shortcut exacerbates the normal problems associated with witness anonymity, while also depriving the accused of the right to confront his accusers in open court, even if he retains the possibility to “question” them at an earlier stage. Nevertheless, a number of states allow the use of hearsay evidence,²¹ and some commentators and institutions have even called for this practice to be extended, perhaps with the introduction of safeguards such as videotaping the original statement. The Committee of Ministers of the Council of Europe has encouraged the use of statements given during the preliminary phase of the procedure as evidence in court when it is not possible for witnesses to appear before the court or when appearing in court might result in great and actual danger to the witnesses/collaborators of justice or to people close to them. Consequently, pre-trial statements should be regarded as valid evidence if the parties have, or have had, the chance to participate in the examination and interrogate and/or cross-examine the witness and to discuss the contents of the statement during the procedure.²²

21. See, for example, s. 96 of the German Code of Criminal Procedure, which allows under certain circumstances for hearsay evidence to be taken from those who have questioned a witness who is to remain anonymous, and used as substantive evidence at trial. See Fyfe and Sheptycki, p. 343. ↩
22. See Council of Europe, Committee of Ministers, loc. cit. n. 48, at paras. 5 and 17. ↩

The use of anonymous hearsay evidence in trial is thus extremely controversial. While it undoubtedly provides a high level of protection to the witness, it will most often do so at unacceptable cost to the fair trial rights of the accused, as the jurisprudence of the European Court on this matter suggests. It is beyond the scope of this report to evaluate when or whether it might be an acceptable practice.

One of the most common, and cost-effective, mechanisms for operationalizing anonymity orders is the erection of a screen between the witness testifying and the defendant. In many countries, the use of screens for witnesses is considered an unproblematic — if usually exceptional — measure.²³ To protect anonymity, a common practice is for the witness in question to be assigned a pseudonym, which is then used for all trial purposes and public records. If necessary to protect the witness’s identity, voice-distorting technology is also used. Where the decision has been made to keep a witness’s identity secret from the accused, the use of screens provides one way of balancing the right to fair trial with the exigencies of witness protection that is significantly more robust than the use of anonymous hearsay evidence in that it allows for greater cross-examination. The risk that inevitably accompanies courtroom appearances, however, is that a witness’s identity will be disclosed.

23. It is included, for example, in section 13G(1)(b) of the New Zealand Evidence Act of 1908. With respect to the United Kingdom, see also Foster [1995] Crim.L.R. 333. The requisite balancing was also discussed in some detail in Donaughy Re Application for Judicial Review (2002) NICA (8.5.2002), where police witnesses testifying in the Bloody Sunday investigations were allowed to do so from behind screens. ↩

Some of the high risks of identity disclosure associated with courtroom appearances can be lessened by giving testimony through video links. The ICTY has used this option to protect witnesses. Rule 75B(i) (c) of its Rules of Evidence and Procedure states that the Chamber can hold in camera proceedings to decide whether to allow, inter alia, the “giving of testimony through image- or voice-altering devices or closed circuit television”.²⁴ The use of videoconferencing (that is, witnesses testifying via video-link from locations far away from The Hague) has also been allowed by the Tribunal due to the “extraordinary circumstances” in which it operates, provided that the testimony is so important that it would be unfair to proceed without it, and the witness in question is unwilling or unable to come to the seat of the ICTY. The Trial Chamber appointed a presiding officer who was to be present when the witness was testifying, to ensure that it was done freely, to identify the witness, and to administer the oath; moreover, only he and some technical staff were permitted to be present. Lastly, the witnesses had to be able to see the judge, accused and questioner on a monitor, and had to be seen by them on theirs.

24. Such measures are granted by the ICTY almost as a matter of course; see Prosecutor v. Sikirica et al., IT-95-8 (2001) (order on request for protective measures), in which protective measures of this sort were ordered for 23 witnesses. ↩

By the same token, these technological solutions come with problems of their own: Quite apart from creating a new layer of technical issues and costs, the witness is not subject to the symbolism of the courtroom, nor fully to the solemnity of its procedures. This in itself can be viewed as not entirely fair to the accused. It is for these reasons that the ICTY in Prosecutor v. Dusko Tadić saw fit to reaffirm the basic principle that witnesses should be present in the courtroom, and to extend the exceptional measure of videoconferencing only where there were good reasons as to why that was not possible.

Another technique which has been used to reduce the impact of such witness protecting measures on the rights of the defendant is the use of “special counsel” for the accused. This involves a court-appointed lawyer who represents the interests of the accused in any context in which witness protection measures might lessen the effectiveness of his own counsel. The special counsel thus acts in something like an amicus curiae capacity, representing the interests of the accused but with no legal responsibility to him. In cases in which full anonymity has been ordered, then, the special counsel could be provided with full details of the witness’s identity and pose questions based thereon, without any obligation to make these known to the defence. This approach has, for example, been adopted by the United Kingdom in response to certain European Court judgements.²⁵

25. See Special Judgment on Appeals Commission Act, 1997, and the Northern Ireland Act, 1998. For a discussion of this issue, see also 1997 New Zealand Law Commission report on Witness Anonymity, paras. 46-47. ↩

A final option to which reference might be made is for judges to consider making greater use of pre-trial detention involving police defendants who are deemed likely to seek to intimidate witnesses. It is common in Uruguay for accused police officers to be detained pending trial. While a strong rule of this kind would raise due process concerns, an approach which signals a willingness to consider detention if obstruction of justice problems arise might be warranted.

Whatever measures are introduced within the criminal justice system to lessen the risk posed to witnesses to extrajudicial executions by state agents, it will generally also be necessary to adopt a formal witness protection programme for some cases. This section considers questions of formal structure, criteria and procedures for both admission to such programmes and termination of protection, and the basic design of witness relocation plans. In addition, consideration is given to some of the challenges specific to witness protection in cases involving human rights abuse by state agents and to innovative responses that states have made in such contexts.

The structure of a witness protection programme

One of the most well-established witness protection schemes is the United States Federal Witness Security Program. In many respects it can be viewed as “paradigmatic”, having served as a model for similar arrangements in other countries. It has been suggested that a comprehensive witness protection model must include: (a) an organizing committee, composed of policymakers from key stakeholders; (b) an operational team, often composed of officials of those institutions involved in the day-to-day running of the programme; (c) a programme administrator; (d) case investigators, or a specially trained law enforcement unit; and (e) designated contact people in all cooperating authorities, agencies and institutions.²⁶

26. See Finn and Healey (see note 20 above), pp. 59-74. ↩

It is important that information regarding witness identities and locations be carefully safeguarded even within such a structure. Thus, for instance, while the operational team may need to weigh the range of protection options available for a particular witness, there is no reason for this information to flow upward to the group dealing with policy questions.

While some countries establish witness protection programmes as units within the police force, this approach is inappropriate for programmes designed to facilitate cases against state agents involved in human rights abuse. One alternative is to appoint officials in key criminal justice system institutions to a committee to administer or oversee the programme. In Belgium, for example, decisions about whom and when to protect are taken by a Witness Protection Board, consisting of public prosecutors, senior police officers and representatives of the Ministries of Justice and the Interior. Another approach is to create a witness protection agency as a separate body, funded by and ultimately responsible to the government, but with decisions on inclusion and exclusion taken by the project director rather than with broader governmental input on the matter, and with law enforcement agencies playing no direct role in the programme. It is worth noting, however, that, if an independent agency is established, it must be given the necessary resources. In some cases, notionally independent agencies have ended up depending on the police force to implement protection measures due to a lack of internal resources.

A formal, institutionalized structure is preferable to an informal, ad hoc one for several reasons. The first is efficiency: a formal structure can involve all key cooperating stakeholders in the planning and execution of protection programmes, thereby minimizing the risk of breakdowns in communication or cooperation, gaps in services to witnesses, and inefficient or ineffective procedures. The second is security: informal procedures involving many agencies (such as housing applications, welfare benefit transfers, etc.) are unlikely to have the necessary bureaucratic safeguards in place to ensure that the new locations or identities of protected witnesses are not too readily disclosed. The third is consistency: ad hoc arrangements are extremely sensitive to changes in personnel at the various cooperating agencies. The fourth is communication: formal, structured programmes assign witnesses with a single, constant contact person, who can provide the round-the-clock support and assistance that many witnesses require if they are to testify. It is important that witnesses are able to build up relationships of trust with the programme administrators with whom they deal, and this is not possible if the latter keep changing. The final reason for preferring a formal, structured programme is evaluation. Frequent and well-planned evaluations are critical to the success of any witness protection effort, allowing administrators to fine-tune the programme and prevent any errors that may have been made from happening again.²⁷

27. Ibid., pp. 59-60. ↩

With respect to criteria for admission to the programme, key considerations include: (a) the importance of the case in terms of ending a cycle of impunity for human rights abuse; (b) the importance of the witness’s testimony to the case; (c) the level of threat to the witness; (d) the suitability of the witness for the programme, including whether he or she is genuinely willing to relocate and break ties with family and friends; and (e) the availability and adequacy of less onerous forms of protection.²⁸

28. Similar criteria may be found in the legislation of a number of countries. See e.g. art. 6 of the Witness Protection Programme Act of Canada, 1996. ↩

It is equally important to develop a clear set of criteria as to when the programme participation of a witness or family member may be terminated. Most jurisdictions provide for termination if the participant breaks any of the rules of conduct that have been agreed upon (on which more below), if they in some way threaten the security of the programme itself, or if the circumstances which necessitated the original provision of protection have ceased to exist. They also always provide for some sort of review by an official or body other than the programme director or administrator.²⁹ The provisions of the Witness Protection Act of South Africa are perhaps among the clearest and most detailed in this regard, stating that the Director may discharge any protected person from protection if he is of the opinion that:
(a) the safety of the person is no longer threatened;
(b) satisfactory alternative arrangements have been made for the protection of the person;
(c) the person has failed to comply with any obligations imposed upon him or her by or under this Act or the protection agreement;
(d) the witness, in making application for placement under protection, wilfully furnished false or misleading information ...;
(e) the person refuses or fails to enter into a protection agreement when he or she is required to do so ...;
(f) the behaviour of the person has endangered or may endanger the safety of any protected person or the integrity of a witness protection programme ...; or
(g) the person has wilfully caused serious damage to the place of safety where he or she is protected, or to any property in or at such place of safety.³⁰

29. See, for ex., the Victoria State Witness Protection Act 1991, sect. 16. ↩
30. Witness Protection Act of 1998, Article 13. ↩

The legislation also provides for the right to request a review of any decision to terminate, to be made by the relevant Government Minister.³¹

31. Witness Protection Act of 1998, Article 14. ↩

Establishing mechanisms for inter-agency cooperation is essential to an effective witness protection programme. Such programmes need to provide much more than simply physical protection for those in their care; a whole host of other issues inevitably arise, from health to housing and other welfare benefits and beyond. These difficulties are, moreover, compounded when witnesses are provided with new and secret identities or addresses, and, as the number of different agencies and institutions involved increases, so do the chances for either inefficient bureaucratic procedures, opportunities to exploit the system or security lapses. The risk of all three can be minimized by employing certain good practices including:
(a) Establishing a contact person within each agency who is in a position to take action when a request is made. Again, the creation and maintenance of relationships of mutual trust between individuals in the various government agencies and officials of the witness protection programme can be crucial in ensuring that requests for assistance are dealt with in a timely and efficient manner;
(b) Gaining support from the top. It is important to ensure that the head of each agency is both aware and approving of the relationship between the witness protection officials and their contact person within the relevant agency, and of the actions taken by the latter at the request of the former;
(c) Developing inter-agency memorandums of understanding. These written agreements between agencies can be extremely useful in ensuring the clarity, consistency and efficiency of the inter-agency relationship. The reasons for using these are many and varied: they create stability despite personnel changes in either agency; making a commitment in writing makes agencies less likely to attempt to evade responsibility, but it also makes it less likely that they will be asked to do more than they should; and a written document serves to reduce uncertainties surrounding the role and responsibilities of each party. The contents of the memorandums will vary from context to context; however, at a minimum, each should specify the services each agency will provide, the staff and funding they will make available, and the allowable expenses or services.³²

32. Finn and Healey (see note 20), pp. 71-74; see also Kelly Dedel, Witness Intimidation (U.S. Department of Justice, Community Oriented Policing Services, July 2006), pp. 18-19. ↩

Protection measures short of relocation

Within the context of a formal witness protection programme, there can be a range of possible protective measures employed on the basis of a case-specific risk assessment. In some countries, the measure of first resort has been protective incarceration. Witnesses are placed in what amounts to a cell inside a police station, unused prison, or other security forces establishment. Except as an extremely temporary measure, this kind of approach must be avoided. Witnesses quite reasonably can seldom accept their own detention as a condition for participating in what will generally be a lengthy trial. Nevertheless, it is not always necessary or advisable to escalate directly to witness relocation or identity change. Sometimes it may be possible to employ lower level physical protection measures. One possibility is the provision of a “rapid response alarm”, connected directly to the local police station. This allows officers to arrive quickly if any threat materializes, without having the expense of providing round-the-clock protection.³³ Another possibility may be the installation of locks, grates, security alarms and outdoor lighting for the witness’s house,³⁴ or increasing patrols in the area. On occasion, police officers in some jurisdictions can assist by taking at-risk witnesses and their families to and from work and school for a short time. Lastly, it is common practice for police to send vulnerable witnesses to live with out-of-town relatives (as a cheap alternative to making complex and costly relocation arrangements within the witness protection programme).³⁵ Many of these measures will, however, be generally inadequate when witnesses fear retaliation from the police or other security forces.

33. Maynard (see note 6 above), p. 4. ↩
34. Dedel (see note 32), pp. 22-23. ↩
35. Finn and Healey (see note 27), p. 29. ↩

Relocation is the central mechanism in effective witness protection programmes and is generally viewed as the most effective way of protecting high-risk witnesses from intimidation, threats and violence. It is, however, also often expensive and time-consuming, and it requires a high degree of inter-agency cooperation and planning if it is to be carried out effectively.³⁶ There are broadly three different types of relocation: emergency, short-term, and long-term/permanent.³⁷

36. Dedel (see note 32), p. 27. ↩
37. Finn and Healey (see note 27), pp. 23-38. ↩

Emergency relocation is used when a threat is imminent and often requires an expedited initial process in terms of acceptance into the witness protection programme, with a more detailed evaluation to take place when the danger has passed. The accommodation used is very often a hotel (which is very expensive); however, it can also be a police station or another public building designated for that purpose. Such protective incarceration should, however, never last longer than a few days or weeks, at most.

Short-term relocation is used when the witness remains at risk for longer periods. It is in this context that the practice of sending witnesses to live with family or friends out of town is most popular, as by far the most cost-effective measure; it also provides a source of emotional support and means that a number of risks attendant to relocation measures, in particular boredom and subsequent returns to visit friends and family, are minimized.³⁸ However, depending on the structure of a society and the character of the threat, this may not provide adequate protection. Indeed, in some circumstances, it will only increase the danger to family members. Other possibilities include social housing or rental property. Here already, however, logistical difficulties begin to arise: it may be necessary to transfer schools; the witness will in all likelihood no longer be able to work; it may also be necessary to make arrangements for the transfer of social benefits to the other jurisdiction. As always, the more people and agencies are involved in the process, the higher the risk of disclosure of the whereabouts of the witness, accidental or otherwise.

38. Finn and Healey (see note 27), p. 29. ↩

Long-term/permanent relocation is most commonly used in cases in which the threat of violent retaliation does not end even with the conviction of the defendant. This has often been true in major gang-related or organized crime cases, but the same problem exists when the police or security forces continue to commit abuses with impunity and may retaliate on behalf of their convicted colleague. Permanent relocation need not be significantly more costly than its short-term counterpart. The major outlays mostly come at the beginning in any event, such as housing costs, and subsistence payments until either social security benefits can be transferred or the witness finds employment. 35 Finn and Healey (see note 27), p. 29. 36 Dedel (see note 32), p. 27. 37 Finn and Healey (see note 27), pp. 23-38. 38 Finn and Healey (see note 27), p. 29.

Innovative approaches to relocation

There are some innovative approaches to relocation that may be useful in meeting the special challenges of protecting witnesses to human rights abuses perpetrated by state agents. These include involving foreign governments and non-governmental organizations in the implementation of relocation plans.

It is relatively common for non-governmental human rights advocacy organizations to help victims and witnesses on an ad hoc basis. In some cases, such assistance has been transformed into full-fledged witness protection programmes and even into joint arrangements between government and non-governmental organizations. For example, in Brazil the witness protection programme began as a project of the non-governmental organization Gabinete de Assessoria Jurídica às Organizações Populares (GAJOP) and was subsequently developed into a programme that involves both government agencies and a number of non-governmental organization partners.³⁹ Today, a committee that includes judges, prosecutors, and others, provides policy direction and makes final decisions on the admission and expulsion of witnesses while day-to-day operations are conducted by the state’s secretariat for justice in tandem with a non-governmental organization.⁴⁰ The non-governmental organization receives government funds to relocate witnesses and help them integrate into a new community. This innovative structure, in which government officials are not actually informed of the witness’s location, has provided witnesses to crimes committed by government agents a much higher level of protection than most systems that rely solely on the government to provide protection. However, some of the non-governmental organizations providing protection services to witnesses reported dissatisfaction with the structure of the programme and questioned the long-term viability of a programme that relies so extensively on non-governmental organization implementing partners. Non-governmental organization involvement in witness protection can be invaluable in overcoming the difficulties caused by a severe lack of trust in state institutions, and in law enforcement agencies in particular.

39. The involvement of non-governmental organizations also features prominently in the history of witness protection in South Africa. ↩
40. Law No. 9.807 (13 July 1999); Decree No. 3.518 (20 June 2000). ↩

It is important to note, however, that even a strong and respected non-governmental organization will find protection difficult or impossible without corresponding government action. The process of investigation, prosecution, and trial will still require contact between witnesses and government officials, and reforms to minimize the risks posed by these contacts will still be necessary. Moreover, it is impossible for non-governmental organizations to arrange new identities for witnesses without the cooperation of numerous government agencies. Finally, the risks posed to non-governmental organization staff implementing such a programme may exceed even those posed to government witness protection personnel.

When secure and trusted relocation within a state’s territory is not feasible, whether due to limited government capacity or to the pervasiveness of the threat, relocation to another country may be considered. South Africa is one state that has successfully resorted to international cooperation on witness protection to enable the prosecution of human rights cases. For example, in the trial of the commander of a police hit squad, Eugene de Kock, three witnesses who were police officers themselves and feared intimidation from colleagues, were sent to Denmark for 18 months. This type of cooperation is already common, on an informal basis, between the heads of the European witness protection programmes, and a best practice survey by the Council of Europe has called for more formalized action along these lines.⁴¹ It is worth considering whether countries with established witness protection programmes could more routinely agree to relocate witnesses to their territory in order to assist countries attempting to successfully prosecute state agents so as to break with a pattern of impunity. Any such arrangement would, of course, need to take into account how witness participation in investigations and trials could be ensured. Either secure transport involving both governments would need to be provided or the provision of testimony through videoconferencing would need to be authorized.

41. Council of Europe Witness Protection Best Practice Survey, 1999, p. 25. ↩

When extrajudicial executions are committed by military personnel it usually falls to the national system of military justice to investigate, prosecute and punish. Yet, historically, the human rights track record of military justice has been dismal. Commanders have routinely used their power to mete out military justice to absolve their personnel of responsibility for brutal abuses against civilians — even as they have harshly punished conscripts and enlisted men for minor breaches of military discipline. At its best, military justice has been a separate and inferior system of justice. At its worst, it has provided a pretext for impunity. In too many countries around the world military justice systems continue to have little in common with human rights and are thus ineffectual in responding to extrajudicial executions.

On a positive note, an increasing number of States have adopted far-reaching reforms designed to subject soldiers to criminal justice systems consistent with international human rights norms. Procedures and institutional structures have been reformed and the relevance of the civilian justice system expanded. This section reviews the experience of several countries with widely different legal traditions which have instituted major reforms. The purpose of the review is to illustrate the problems identified in the previous systems, and to show that fundamental reform is both necessary and feasible.

Two caveats are applicable to the following analysis. First, it is based on published analyses rather than empirical surveys.⁴² There may thus be significant discrepancies between theory and practice which are not reflected here. Second, the focus is on procedures relating to violations of the right to life and not on other offences.

42. Particularly helpful resources included Justicia Militar, Códigos Disciplinarios y Reglamentos Generales Internos, Informe Final (eds. Gustavo Fabián Castro & Dolores Bermeo Lara) (2008); Arne Willy Dahl, “International Trends in Military Justice” (January 2008) at <http://home.scarlet.be/~ismllw/index_UK.htm>; Victor Hansen, “Changes in Modern Military Codes and the Role of the Military Commander: What Should the United States Learn from this Revolution?”, 16 Tulane Journal of International & Comparative Law (2008) 419; Peter Rowe, The Impact of Human Rights Law on Armed Forces (2006); Federico Andreu-Guzmán, Military Jurisdiction and International Law: Military Courts and Gross Human Rights Violations (2004); European Military Law Systems (ed. Georg Nolte) (2003); Eugene R. Fidell, “A World-Wide Perspective on Change in Military Justice”, 48 Air Force Law Review (2000) 195. ↩

The obligation to effectively investigate, prosecute, and punish violations of the right to life in situations of armed conflict has been a consistent theme of reports under this mandate.⁴³ The most systematic survey of the implications of international law in this area is contained in the report submitted by the Special Rapporteur of the Sub-Commission on the Promotion and Protection of Human Rights,⁴⁴ Emmanuel Decaux, which included draft principles governing the administration of justice through military tribunals. These principles state that “the jurisdiction of military courts should be limited to offences of a strictly military nature”.⁴⁵ That jurisdiction “should be set aside in favour of the jurisdiction of the ordinary courts to conduct inquiries into serious human rights violations such as extrajudicial executions, enforced disappearances and torture, and to prosecute and try persons accused of such crimes”.⁴⁶

43. See, for ex., E/CN.4/1995/61, para. 93; E/CN.4/1999/39, para. 67; and E/CN.2006/53, paras. 33-43. ↩
44. E/CN.4/2006/58. ↩
45. Ibid, paras. 29-31 (Principle No. 8). ↩
46. Ibid., paras. 32-35 (Principle No. 9). ↩

The principles also state that “the organization and operation of military courts should fully ensure the right of everyone to a competent, independent and impartial tribunal at every stage of legal proceedings from initial investigation to trial. The persons selected to perform the functions of judges in military courts must display integrity and competence and show proof of the necessary legal training and qualifications. Military judges should have a status guaranteeing their independence and impartiality, in particular vis-à-vis the military hierarchy”.⁴⁷

47. Ibid., paras. 45-48 (Principle No. 13). ↩

Violations of these norms affect the rights of both the accused and the victims. The Inter-American Court of Human Rights has emphasized the “close connection”⁴⁸ among these rights, has developed a unified jurisprudence looking at questions of independence and jurisdiction from both perspectives, and has gone so far as to refer to the rights of victims to due process.⁴⁹

48. See Vargas-Areco v. Paraguay (2006), para. 73. ↩
49. Las Palmeras Case (2001), para. 54. ↩

The United Kingdom provides an important case study in rapidly transforming a military justice system of ad hoc courts-martial dominated by military commanders, of the type that still exists in many other States today, into one overseen by a standing court and independent and professional prosecutors and judges.

While national institutions drove the reform process, the European Court of Human Rights played a catalytic role.⁵⁰ In Findlay v. the United Kingdom, the European Court ruled that the British system of military justice was incompatible 43 See, for ex., E/CN.4/1995/61, para. 93; E/CN.4/1999/39, para. 67; and E/CN.2006/53, paras. 33-43. 44 E/CN.4/2006/58. 45 Ibid, paras. 29-31 (Principle No. 8). 46 Ibid., paras. 32-35 (Principle No. 9). 47 Ibid., paras. 45-48 (Principle No. 13). 48 See Vargas-Areco v. Paraguay (2006), para. 73. 49 Las Palmeras Case (2001), para. 54. 50 M. Oakes, The Armed Forces Bill, House of Commons Research Paper 01/03 (8 January 2001), p. 12. with the requirement of the European Convention on Human Rights of an “independent and impartial tribunal”.⁵¹ At that time, the British system conferred broad authority on the “convening officer”, who was an officer with command authority over the accused.⁵² Among other responsibilities, the convening officer would decide which charges would be brought, choose the members of the courtmartial (who could be his subordinates), and appoint the prosecuting officer.⁵³ The convening officer could also “discontinue the court-martial either before or during the trial”.⁵⁴ The convening officer also generally played the role of the “confirming officer” with the power to “withhold confirmation or substitute, postpone or remit in whole or in part any sentence”.⁵⁵ Thus, those responsible for judging the accused would be appointed by and often be subordinate to the accused’s commanding officer. Moreover, the ultimate authority both to prosecute and to sentence rested with the commander. The role of the Judge Advocate General was limited to providing advice to the convening officer, the members of the court-martial, and the confirming officer.⁵⁶

50. M. Oakes, The Armed Forces Bill, House of Commons Research Paper 01/03 (8 January 2001), p. 12. ↩
51. European Convention on Human Rights, art. 6 (1); European Court of Human Rights, Findlay v. the United Kingdom, judgement of 25 February 1997, paras. 68-80. ↩
52. Findlay (see note 51 above), para. 36. ↩
53. Ibid., paras. 34-37. ↩
54. Ibid., para. 40. ↩
55. Ibid., para. 48. ↩
56. Ibid., paras. 42-45. ↩

The far-reaching reforms in the United Kingdom began in 1996.⁵⁷ While the new system leaves commanders with considerable discretion in conducting summary hearings for various disciplinary breaches,⁵⁸ it shifts most authority to professional investigators and prosecutors in relation to serious crimes, such as murder. In such cases, the commanding officer is obligated to bring the crime to the attention of the military police.⁵⁹ The police must investigate and, should the evidence warrant, refer the case to the Director of Service Prosecutions (DSP).⁶⁰ It is the Director of Service Prosecutions who then has the authority to direct that charges be brought.⁶¹ (The powers of the Director of Service Prosecutions may be delegated to legally qualified “prosecuting officers” appointed by him.⁶²) The Court Martial, a standing body, itself is composed of a judge advocate and a number of “lay members”.⁶³ The judge advocate is appointed by the Judge Advocate General, and the lay members, who are officers or warrant officers, are appointed by the “court administration officer”.63 The judge advocate makes binding rulings on questions of law and procedure, the lay members make the finding as to guilt, and the lay members and the judge advocate together decide on the sentence.⁶⁴ If someone is convicted by the Court Martial, he or she may appeal the conviction and sentence to the Court Martial Appeal Court.⁶⁵ The Attorney General may also refer a case to this appeal court if he concludes that the sentence imposed by the Court Martial was “unduly lenient”.⁶⁶ Points of law may also be referred to the country’s highest court.⁶⁷

57. Armed Forces Act 1996; Armed Forces Discipline Act 2000; Armed Forces Act 2001; Armed Forces Act 2006. With respect to the phased implementation of reforms, see Armed Forces Act 2006, para. 383, and the statutory instruments made by the Secretary of State, including especially those made on 10 May 2007, 8 October 2007 and 24 June 2008. ↩
58. The Bill’s provisions retain the Commanding Officer at the centre of the summary dealing process, which represents more than 95 per cent of Service discipline cases. House of Commons, Session 1995-96, Special Report from the Select Committee on the Armed Forces Bill, HC 828-I (30 April 1996), para. 43. ↩
59. Ibid., paras. 113-115. ↩
60. Ibid., para. 116. ↩
61. Ibid., paras. 113-122. ↩
62. Ibid., para. 365. ↩
63. Ibid., para. 155. ↩
64. Ibid., paras. 159-160. ↩
65. Courts-Martial (Appeals) Act 1968, para. 8, as amended by the Armed Forces Act 2006, Schedule 8. The court itself is composed of judges and other persons with legal experience appointed by the Lord Chief Justice and the Lord Chancellor. ↩
66. Ibid., Armed Forces Act 2006, para. 273. ↩
67. Ibid., para. 274. ↩

Decision-making authority has, thus, been shifted from commanders to the Director of Service Prosecutions, Judge Advocate General, and court administration officers. In varying manners and to varying degrees, provisions have been made to ensure that these institutions are independent from each other, the chain of command, and the executive branch. The Director of Service Prosecutions is legally qualified and appointed by the Queen.⁶⁸ The Judge Advocate General is appointed by the Queen on the recommendation of the Lord Chancellor.⁶⁹ The court administration officer of the Court Martial is appointed by the Defence Council, which is chaired by the Secretary of Defence.⁷⁰

68. Ibid., para. 364. ↩
69. Courts-Martial (Appeals) Act 1951, Ch. 46, para. 29, as amended by the Armed Forces Act 2006, Schedule 16. ↩
70. Armed Forces Act 2006, para. 363. ↩

There are also lessons to be learned from a reform process that has made periodic reviews of the military justice system routine.⁷¹ In the United Kingdom, as a matter of constitutional law, statutes establishing the military justice system will lapse unless parliament enacts new legislation every five years. In practice, this has meant that both the Ministry of Defence and a special committee established by the lower house of parliament systematically review the existing military justice system in light of the experience of the intervening years, changed circumstances and human rights norms.⁷²

71. See also E/CN.4/2006/58, paras. 64-66 (Principle No. 20). ↩
72. M. Oakes [(see note 50 above), pp. 10-14] discusses this review process. ↩

Reforms adopted in Colombia provide a case study in transferring jurisdiction over serious human rights violations, such as extrajudicial executions, from military jurisdiction to that of ordinary courts.

The reform process in Colombia has been driven by efforts to interpret and apply the provisions of the Constitution of 1991, including its provision that crimes committed by members of the military and police are subject to military jurisdiction when they are “in active service” and the conduct was “in relation to that service”.⁷³ In 1997 the Constitutional Court held that this must be construed narrowly to require a “proximate and direct” link between the crime and some broader mission or task that falls within the constitutionally defined role of the security forces.74 It 64 Ibid., paras. 159-160. 65 Courts-Martial (Appeals) Act 1968, para. 8, as amended by the Armed Forces Act 2006, Schedule 8. The court itself is composed of judges and other persons with legal experience appointed by the Lord Chief Justice and the Lord Chancellor. 66 Ibid., Armed Forces Act 2006, para. 273. 67 Ibid., para. 274. 68 Ibid., para. 364. 69 Courts-Martial (Appeals) Act 1951, Ch. 46, para. 29, as amended by the Armed Forces Act 2006, Schedule 16. 70 Armed Forces Act 2006, para. 363. 71 See also E/CN.4/2006/58, paras. 64-66 (Principle No. 20). 72 M. Oakes [(see note 50 above), pp. 10-14] discusses this review process. 73 See Article 221. 74 See Sentencia C-358-97 (5 August 1997). noted that a broader interpretation risked transforming military jurisdiction into the personal privilege of a separate military class. It held that there are acts, such as crimes against humanity, that are “so flagrantly at odds with the constitutional function of the [military and police forces] that their commission alone breaks any functional link between the agent and the service” regardless of the context in which they were committed and that, thus, must be tried by a court of general jurisdiction.74 This reflected the court’s earlier judgement that international obligations prevented the concept of due obedience from being interpreted so as to extinguish criminal liability for human rights violations committed by soldiers.⁷⁵

73. See Article 221. ↩
75. Sentencia C-578/95 (4 December 1995). ↩

Judgments of the Inter-American Court of Human Rights have reinforced this approach. The Court has held that “the penal military jurisdiction shall have a restrictive and exceptional scope” and may only extend to offences “against legally protected interests of military order”.⁷⁶ The American Convention on Human Rights requirement that crimes be tried by a “competent” tribunal would be violated if ordinary crimes were to be tried under military jurisdiction.⁷⁷ The Court has ruled against the use of military jurisdiction in various cases, including one involving extrajudicial executions by police during a counterinsurgency operation⁷⁸ and a case in which a state’s military suppressed prison riots, resulting in 111 deaths.⁷⁹

76. Durand and Ugarte v. Peru, judgement of 16 August 2000, para. 117. ↩
77. Ibid., paras. 113-131. ↩
78. Las Palmeras, paras. 2, 32-47. ↩
79. Durand and Ugarte (see note 76 above), para. 59. ↩

In 1999, the Congress enacted a new military penal code which reiterated that military jurisdiction is contingent upon a link between the crime and a proper military function and expressly barred military jurisdiction over the crimes of torture, genocide, and enforced disappearance,⁸⁰ a list which is not considered exhaustive.⁸¹ Similarly, in 2000, the Congress adopted a new general penal code, which contains a section on “crimes against persons and goods protected by international humanitarian law”,⁸² which includes “homicide of a protected person”.⁸³

80. Código Penal Militar, Law 522 of 1999 (13 August 1999), arts. 1-3. ↩
81. Sentencia C-878/00 (12 July 2000). ↩
82. Código Penal, Law 599 of 2000 (24 July 2000), Title 2. ↩
83. Código Penal, art. 135. ↩

Today, an extrajudicial execution by a Colombian soldier may be characterized as an offence under the ordinary penal code. This offence would then be prosecuted by the Attorney General’s Office and tried before a court of general jurisdiction. Disputes regarding whether jurisdiction lies with the military justice system or the regular courts are resolved by the Supreme Judicial Council.⁸⁴

84. Constitution, art. 256(6). The Supreme Judicial Council is composed of judges serving eight year terms (Constitution, art. 254). For a discussion of jurisprudence on this issue, see Military Jurisdiction and International Law, pp. 241-247. ↩

Since 1991, the Netherlands has had neither military courts nor courtsmartial,⁸⁵ although commanders still play a large role in disciplinary cases.86 But 75 Sentencia C-578/95 (4 December 1995). 76 Durand and Ugarte v. Peru, judgement of 16 August 2000, para. 117. 77 Ibid., paras. 113-131. 78 Las Palmeras, paras. 2, 32-47. 79 Durand and Ugarte (see note 76 above), para. 59. 80 Código Penal Militar, Law 522 of 1999 (13 August 1999), arts. 1-3. 81 Sentencia C-878/00 (12 July 2000). 82 Código Penal, Law 599 of 2000 (24 July 2000), Title 2. 83 Código Penal, art. 135. 84 Constitution, art. 256(6). The Supreme Judicial Council is composed of judges serving eight year terms (Constitution, art. 254). For a discussion of jurisprudence on this issue, see Military Jurisdiction and International Law, pp. 241-247. 85 For this legislation and its background, see Military Jurisdiction and International Law, pp. 293-294. cases implicating soldiers, including those arising in armed conflict situations, are dealt with under the same system of trial and appellate review that is used for ordinary criminal cases implicating civilians.⁸⁷ However, cases involving military personnel in both the district court and the court of appeal use a special military chamber comprising two civilian judges and one military member.⁸⁸ The cases are prosecuted by a regular, civilian prosecutor attached to the Office of the Public Prosecutor (Openbaar Ministerie).⁸⁹

85. For this legislation and its background, see Military Jurisdiction and International Law, pp. 293-294. ↩
87. A.F.M. Brenninkmeijer, “Judicial Organization”, Introduction to Dutch Law, pp. 56-57. ↩
88. L. Besselink (see note 86 above), p. 631. ↩
89. Ibid., p. 632. ↩

Most rules of substantive and procedural law are provided by the general criminal code and code of criminal procedure, but there are some military-specific offences and rules included in the Military Criminal Code and Act of Military Criminal Procedure.⁹⁰ These include provisions to increase sentences and provide additional defences to criminal liability when an offence takes place in an armed conflict.⁹¹ There is also the Laws of War Act, which criminalizes violations of “the laws and customs of war”.⁹²

90. Ibid., pp. 630-631. ↩
91. D-J Dieben & T. Dieben, When does War become Crime? Aspects of the Criminal Case against Eric O. (2005), pp. 21-22; P. C. Tange, “Netherlands State Practice for the Parliamentary Year 2005-2006”, Netherlands Yearbook of International Law (2007), pp. 307-309, 342-349 (excerpting letters from Minister of Defence to members of the parliament concerning deployments in Afghanistan and other contexts). ↩
92. L. Besselink (see note 86), pp. 632-636. ↩

Procedures and institutions designed to protect witnesses, including formal witness protection programmes, must be central to any overall effort to punish those responsible for extrajudicial executions. Yet, many countries in which such executions are common either have no programme worthy of the name, or one that functions very poorly.

The lack of serious witness protection programmes is certainly due in part to resource constraints. But much more importantly, it is a symptom of the absence of political will to ensure accountability for extrajudicial executions, especially when those suspected are police or military officers.

While the focus of the analysis above is on the need for carefully designed institutional and other arrangements, institutional design per se can never overcome a lack of political will.

The international community has recognized the importance of witness protection in contexts like the international criminal tribunals, but has failed to provide the financial, technical and political support needed to develop effective programmes at the national level in most situations where action is urgently needed. 86 L. Besselink, “Military Law in the Netherlands”, European Military Law Systems, pp. 624-628. 87 A.F.M. Brenninkmeijer, “Judicial Organization”, Introduction to Dutch Law, pp. 56-57. 88 L. Besselink (see note 86 above), p. 631. 89 Ibid., p. 632. 90 Ibid., pp. 630-631. 91 D-J Dieben & T. Dieben, When does War become Crime? Aspects of the Criminal Case against Eric O. (2005), pp. 21-22; P. C. Tange, “Netherlands State Practice for the Parliamentary Year 2005-2006”, Netherlands Yearbook of International Law (2007), pp. 307-309, 342-349 (excerpting letters from Minister of Defence to members of the parliament concerning deployments in Afghanistan and other contexts). 92 L. Besselink (see note 86), pp. 632-636.

The Office of the United Nations High Commissioner for Human Rights should develop policy tools designed to encourage and facilitate greater attention to witness protection in national level programmes to combat impunity for killings and other crimes.

In too many countries around the world military justice systems are incompatible with human rights obligations. The impunity that they foster also makes them ineffectual in responding to extrajudicial executions.

Governments should periodically review their military justice systems in light of human rights norms and trends in state practice.

The General Assembly should call upon all States to report on the extent to which their existing arrangements comply with human rights standards.