1  The present entry draws upon a previous paper by the author (Roberts, 2021).

2  This entry briefly outlines issues pertaining to proof at what are sometimes termed quasi-judicial human rights bodies (quasi-judicial bodies). Issues pertaining to proof have not been dealt with particularly clearly or extensively by such human rights bodies, which have also displayed a certain degree of variation among themselves. The following account attempts to trace what appear to be common principles and approaches with attention to several bodies, as well as highlighting certain forms of cross-tribunal variations, where they occur.

3  Particular attention is paid to the jurisprudence and practice of nine bodies: the Inter-American Commission on Human Rights (IACommHR), the African Commission on Human and Peoples’ Rights (ACommHPR), the Human Rights Committee (‘UNHRC’), the Committee on the Elimination of Racial Discrimination (CERD), the United Nations Committee Against Torture (CAT), the Committee on the Elimination of Discrimination Against Women (CEDAW), the Committee on the Rights of the Child (CRC), the Committee on the Rights of Persons with Disabilities (CRPD), and the Working Group on Arbitrary Detention (‘WGAD’). From among the various human rights bodies that might have been selected, these bodies have been focused on due to their more extensive contemplation of the relevant issues. While the jurisprudence of other bodies, such as the Committee on Economic, Social and Cultural Rights (CESCR), the Committee on the Protection of the Rights of All Migrant Workers and Members of their Families (CMW), the Committee on Enforced Disappearances (CED), and other international and regional bodies (Committees: Human Rights Bodies) is not considered below, the principles and approaches laid out may be understood as generally applicable to those bodies as well.

4  While this entry does not specifically focus on the approach of the European Court of Human Rights (ECtHR) or the Inter-American Court of Human Rights (IACtHR), substantial attention is also paid to the jurisprudence and approach of both, as well as to a lesser extent to the African Court on Human and Peoples’ Rights (ACtHPR). This is due to the fact that those bodies have given more extensive attention to issues of proof; to the fact that the interpretations of the IACtHR and ACtHPR carry special weight relative to the commissions in those regions; and to the fact that the approaches of those bodies carry persuasive weight at all human rights bodies. For these reasons, understanding the standards applied by the regional human rights courts is essential in order to understand the jurisprudential field in which human rights quasi-judicial bodies are operating when confronting issues of proof, though the less judicialized nature of non-fully judicial human rights bodies must be borne in mind as well.

5  As with other adjudication, questions relating to burden of proof and standard of proof are central to the issue of evidential consideration before human rights bodies (Assessment of Evidence). By way of brief general definition, the burden of proof refers to the issue of which party bears the burden of convincing the court, overall or on a particular issue, while the standard of proof refers to the level of certainty the party bearing the burden must produce in the adjudicator in order to prevail in the case.

6  In terms of the burden of proof, in common with other international adjudicatory bodies, human rights bodies may be understood as following, as a default, the rule of onus probandi incumbit actori—that is, that ‘it is the duty of the party which asserts certain facts to establish the existence of such facts’ (Pulp Mills on the River Uruguay, Argentina v Uruguay, 2010, para 162). For its part, the IACtHR has held that ‘the plaintiff should, in principle, undertake the burden of proof regarding the facts connected with its arguments’ (Kawas-Fernandez v Honduras, 2009, para 95). This does not mean that it will be for the claimant to prove every point: while as a default it will be for the claimant to prove the facts they want to assert, it will be for the respondent to establish any facts they argue. This principle has been recognized by the ACommHPR, which has observed ‘[t]he burden to produce evidence in support of an alleged fact lies with the party asserting the fact, as a general principle’ (Ezzat and Enayet (represented by Egyptian Initiative for Personal Rights and INTERIGHTS) v Egypt, 2016, para 171; see also Paúl, 2018, para 14). For its part, the UNHRC has observed that ‘the burden of proof cannot rest solely on the author of the communication’, and the CRC and CAT have observed similarly (Shikhmuradov v Turkmenistan, 2015, para 6.2; see also AM v Switzerland, 2021, para 10.5; X v Netherlands, 2019, para 8.5).

7  This rule is a default only—in practice, circumstances may justify deviation, as explored further below. In addition, to the extent human rights bodies apply such a rule, it seems they do so to a significant degree for convenience sake rather than as a matter of substantive justice. As and where such bodies are able to obtain other information, they will take that into account, leaving concerns as to the burden of proof to one side. The ECtHR has for instance held that ‘the burden of proof is not borne by one or the other party because the Court examines all material before it irrespective of its origin, and because it can, if necessary, obtain material of its own motion’ (Merabishvili v Georgia, 2017, para 311; see also Leach, 2018, para 7; Heri, 2018, para 78). If this principle applies in the context of a human rights court, it should apply all the more in the context of a quasi-judicial human rights body, given the greater procedural flexibility afforded to such bodies by their less-judicialized nature.

8  The ACommHPR has suggested it follows a different approach, placing the burden of proof on the government. The Commission has repeatedly indicated that:

9  It is not clear if the ACommHPR’s approach in practice is always as forceful as this statement indicates. Rather, the statement may be a way of indicating the ACommHPR’s general tendency to reverse the burden in cases falling into the categories of cases in which reversing the burden is justified, discussed below, which historically have made up much of the Commission’s caseload.

10  In terms of the standard of proof, international human rights courts have emphasized that the criminal standard is not appropriate in the human rights context. The IACtHR has held, for instance, that ‘[t]he international protection of human rights should not be confused with criminal justice’ (Velasquez Rodriguez v Honduras, 1988, para 134; see also Application of the Convention on the Prevention and Punishment of the Crime of Genocide, Croatia v Serbia, Judgment (Dissenting Opinion of Judge Cançado Trindade), 2015, para 123). The ECtHR has emphasized that ‘[t]he Court’s role, it should be remembered, is to rule not on criminal guilt or civil liability but on Contracting States’ responsibility under the Convention’ (Mathew v Netherlands, 2006, para 156). The ‘beyond a reasonable doubt’ standard applied in the criminal law context hence seems inappropriate in the context of human rights adjudication (Beyond Reasonable Doubt). While quasi-judicial human rights bodies have not articulated their stance on this issue as clearly, there is nothing in their practice or jurisprudence that suggests a different understanding.

11  In this context, it must be noted that a potential exception to the inappropriateness of the ‘beyond a reasonable doubt’ standard in the human rights context applies at the ECtHR, which has repeatedly indicated that it applies the ‘beyond a reasonable doubt’ standard of proof (see, eg, Mathew v Netherlands, 2006, para 156; El-Masri v Macedonia, 2012, para 151). This is misleading, however—as the ECtHR put it in El-Masri, ‘it has never been [the Court’s] purpose to borrow the approach of the national legal systems that use that standard’ (at para 151; see also Thienel, 2007, 578–79; O’Boyle, 2018, para 42). In short, the use of the phrase ‘beyond a reasonable doubt’ in this context does not apparently mean ‘beyond a reasonable doubt’, but rather appears to be the court’s way of indicating that it will not reach the conclusion that human rights have been violated lightly. The ECtHR’s repeated emphasis that it employs such a standard hence seems to be more a rhetorical move designed to reassure State Party respondents than an accurate representation of the standard of proof it in fact employs.

12  While the ‘beyond a reasonable doubt’ standard is hence inappropriate and generally not employed in the human rights context, the precise standard that is applied is not clear. At times, some international courts have indicated that they will require ‘concrete and convincing evidence’, a standard perhaps similar to the intermediate standard of ‘clear and convincing evidence’ employed in certain cases in the United States (see Rajput, 2021, paras 3–4 and 13). This generally corresponds to human rights courts’ recognition of the seriousness of finding that states have committed human rights violations: as the IACtHR has held, ‘[t]he Court cannot ignore the special seriousness of finding that a State Party to the Convention has carried out or tolerated a practice of disappearances in its territory. This requires the Court to apply a standard of proof which considers the seriousness of the charge’ (Godinez Cruz v Honduras, 1989, para 135). The ECtHR has put it in this way: ‘[t]he Court is … attentive to the seriousness that attaches to a ruling that a Contracting State has violated fundamental rights’ (Mathew v Netherlands, 2006, para 156).

13  To the extent that such a standard applies, it seems to be invoked only in the most serious cases. Though it has not been entirely consistent in such regard, the IACtHR has indicated that it applies a higher standard in relation to cases of widespread human rights violations, and the balance of probabilities in other cases (see Paúl, 2018, paras 21–22; Godinez Cruz v Honduras, 1989, para 135). A similar approach is employed by the International Court of Justice (ICJ), which has indicated that a different standard of proof will be applied depending on the circumstances, with ‘conclusive proof’ required in cases of exceptional gravity, and proof on the ‘balance of probabilities’, or the ‘balance of the evidence’, required in other cases (see Application of the Convention on the Prevention and Punishment of the Crime of Genocide, Bosnia and Herzegovina v Serbia and Montenegro, 2007, para 209; Land, Island and Maritime Frontier Dispute, El Salvador v Honduras: Nicaragua intervening, 1992, para 248; Rajput, 2021, paras 10–12 and 14).

14  This is a complex area, however. Public articulation of standards of proof does not only indicate the standard of proof employed—it can also be used to signal the seriousness with which a particular matter is taken. Just as the ECtHR seems to have referred to the beyond a reasonable doubt standard in order to indicate the seriousness with which it approaches its task, so too, other courts’ references to a higher standard may be meant to indicate that they will not reach findings of violation lightly. Articulated standards of proof are not necessarily the standards of proof employed in practice, moreover.

15  For their part, quasi-judicial human rights bodies have not been particularly clear on the standard of proof they employ. In their opinions and decisions, human rights bodies typically use formulas such as that they are ‘of the view that the facts before [them] disclose’, ‘conclud[e] that the facts as submitted reveal’, or simply state that they ‘consider that’ or are ‘of the view that’, which are formulas that leave the precise standard employed vague (Jallow v Denmark, 2022, para 8; Voronkov v Russian Federation, 2022, para 10.5; X and Y v Switzerland, 2022, paras 7.4, 7.7–7.8, and 7.10; see also Román Jaimes v Mexico, 2022, para 8). For several reasons, it may be presumed human rights bodies are employing the ‘more likely than not’ or ‘on the balance of probabilities’ standard, however: first, because that is the standard of proof most frequently employed in adjudication outside the criminal law context; second, because it is the standard that seems to align best with the language human rights bodies utilize; third, because should human rights bodies employ a specific, heightened standard they should presumably be clear as to the approach they are adopting and exactly what standard they are applying; fourth, because the balance of probabilities standard is the standard that would comply with international judicial bodies’ suggestions that a heightened standard is only appropriate in special circumstances; and fifth, because it is the standard that makes the most sense in the context of the readiness with which human rights bodies reverse the burden. Application of a balance of probabilities standard also accords best with the manner in which human rights bodies on occasion reach findings in accordance with the contentions of a party due to ‘the absence of information’ from the opposing party—an absence that might not lead to the same conclusion were a more demanding standard of proof employed (Mursalov and ors v Azerbaijan, 2022, para 9.7).

16  While neither human rights courts nor other human rights bodies have been particularly clear as to the standard of proof employed, there has been much clearer emphasis on the need for flexibility. The IACtHR has observed that ‘international jurisprudence has recognized the power of the courts to weigh the evidence freely’ and ‘has always avoided a rigid rule regarding the amount of proof necessary to support the judgment’ (Velasquez Rodriguez v Honduras, 1988, paras 127–28). The ECtHR has held that it will ‘adop[t] the conclusions supported by the free evaluation of all evidence, including such inferences as may flow from the facts and the parties’ submissions’ (Merabishvili v Georgia, para 315; see also Nachova and ors v Bulgaria, 2005, para 147; O’Boyle, 2018, paras 9–10). The IACommHR has indicated that, ‘following the international case-law and its own practice, [it] shall apply the system of free weighing of the evidence, which gives it a measure of flexibility as regards determining the amount of evidence needed to support the judgment’ (Extrajudicial Executions and Forced Disappearances v Peru, 2001, para 182). The same approach appears to be employed in practice by other human rights bodies as well.

17  While human rights bodies decide cases with a degree of flexibility, that flexibility is nonetheless bound by various principles. This has been recognized by regional courts: thus the IACtHR has held that ‘international legal proceeding[s] … recognize different burdens of proof, depending upon the nature, character and seriousness of the case’ (Velasquez Rodriguez v Honduras, 1988, paras 127–28), and the ECtHR that ‘the level of persuasion necessary for reaching a particular conclusion and, in this connection, the distribution of the burden of proof, are intrinsically linked to the specificity of the facts, the nature of the allegation made and the Convention right at stake’ (El-Masri v Macedonia, para 151). The ACommHPR, meanwhile, has approvingly cited the ICJ’s holding that ‘the determination of the burden of proof is in reality dependent on the subject-matter and the nature of each dispute … it varies according to type of facts which it is necessary to establish for the purpose of the decision of the case’ (Ezzat and Enayet (represented by Egyptian Initiative for Personal Rights and INTERIGHTS) v Egypt, 2016, para 171, citing Ahmadou Sadio Diallo, Guinea v Democratic Republic of the Congo, 2010, paras 54–55).

18  Once again, other human rights bodies may be understood to comply with these principles as well, though they have not articulated them so clearly. In practice, this flexibility is to a significant extent deployed along the lines laid out below, in the discussion of inferences and reversal of the burden of proof.

19  In addition to employing a loose approach to the burden and standard of proof, human rights bodies tend to employ a broad and flexible approach to the admission of evidence. This is true even at human rights courts. The ECtHR has indicated that it applies ‘no procedural barriers to the admissibility of evidence’ (El-Masri v Macedonia, para 151; see also O’Boyle, 2018, para 6); that it will consider more than just ‘direct proof’ (Merabishvili v Georgia, para 316); that, among other sources of information, it will consider ‘[r]eports or statements by international observers, nongovernmental organisations or the media’, as well as ‘the decisions of other national or international courts’ (Merabishvili v Georgia, para 317; see also O’Boyle, 2018, paras 7–8); and that, where feasible, it will ‘obtain material of [its] own motion’ (Merabishvili v Georgia, para 311; see also Heri, 2018, paras 59–68). The IACtHR employs a similarly flexible approach (Paúl, 2018, paras 6–11, 24, and 31–4).

20  Human rights bodies have not stated their approach as clearly, but appear to typically employ a similarly broad and flexible approach in practice. Based on 76 merits decisions by human rights bodies in 2022 (the complete set of such decisions is accessible through the Office of the United Nations High Commissioner for Human Rights’s Jurisprudence database as of 26 February 2023), the work of Human Rights Watch and/or Amnesty International was referenced in 13 cases—ten times by the CAT (DS v Australia, 2022, para 10.3; TB v Switzerland, 2022, para 3.2; TA v Switzerland, 2022, para 3.11; Berhane v Switzerland, 2022, para 3.1; Laaroussi v Morocco, 2022, paras 2.12 and 5.7; SR v Canada, 2022, para 8.3; AY v Switzerland, 2022, paras 4.14 and 8.10; DM v Switzerland, 2022, para 3.2; A and B v Azerbaijan, 2022, para 8.9; X and Y v Switzerland, 2022, para 7.4) and once each by the UNHRC (AB and ors v Poland, 2022, para 7.3), CEDAW (Matson and ors v Canada, 2022, paras 18.3 and 18.9), and CRC (SK v Denmark, 2022, para 3.3). In six of those cases, the work of such organizations was considered in the context of the committees’ findings (Matson and ors v Canada, 2022, paras 18.3 and 18.9; DS v Australia, 2022, para 10.3; SR v Canada, 2022, para 8.3; A and B v Azerbaijan, para 8.9; X and Y v Switzerland, 2022, para 7.4). Clearly, therefore, human rights bodies are not averse to considering the information contained in the reports of non-governmental organizations.

21  In one of those decisions, AY v Switzerland, the CAT included some explicit observations on the appropriateness of considering the work of human rights organizations. The CAT refused to accept the State Party’s assertion ‘that the refusal of a State to engage with or to accept a fact-finding visit is a reason to dismiss the findings of the report of that fact-finding mission, and that information from individuals who have fled persecution is less reliable’, noting that adoption of such an approach ‘would have a dissuasive effect on States engaging with human rights obligations and reduce transparency and accountability’ (AY v Switzerland, para 8.10). In addition, the CAT noted ‘that, on balance, if a particular source were to be given less weight, it should be any that were to be assessed as having a clearly vested interest in the narrative provided’—meaning in particular ‘information provided by government agents or sources within the country, who might face reprisals for any perceived criticism’ (AY v Switzerland, para 8.10). In contrast, the CAT observed that ‘where sources are drawn widely from civil society and citizens in exile, the information provided is more likely on balance to represent reliable, uncensored first-hand accounts, as persons outside the country who have fled are arguably no longer living under censorship’ (AY v Switzerland, para 8.10). Human rights bodies are not only willing to consider reporting from human rights non-governmental organizations, therefore, but have also recognized that that reporting may be more reliable than state accounts.

22  One significant difference between quasi-judicial human rights bodies and human rights courts in terms of the receipt of information relates to the fact that the former engage in a wider set of activities. While activities vary body to body, human rights bodies in general undertake multiple different forms of engagement with human rights issues pertaining to their mandates. These include: performing regular country reviews; preparing substantive and thematic reports; composing ‘general comments’; conducting sessions and hearings; undertaking country visits; and deploying and receiving reports from special rapporteurs. These activities provide human rights bodies with additional sources of human rights-related information. In contrast to human rights courts, therefore—for which case adjudication is the primary task—consideration of complaints forms only one of a set of interlinked, mutually supportive tasks engaged in by quasi-judicial human rights bodies. Human rights bodies hence come to the task of adjudication equipped with multiple channels for the receipt of information, on top of which they may on occasion rely on their own previous reports, as well as external sources. Human rights bodies have frequently relied on the information obtained through other rights review processes they have undertaken in the course of reaching their conclusions on communications (see, eg, AB v Finland, 2021, para 8.6 n 23; Abdel Jalil Laaroussi v Morocco, 2022, paras 8.4–8.5; SR v Canada, 2022, para 8.3; A and B v Azerbaijan, para 8.6; Lauren Henley v Australia, 2022, para 10.10).

23  Human rights courts and quasi-judicial bodies are generally willing to accept ‘third party submissions’ during the course of complaint proceedings, including in the form of amicus submissions. This is true for human rights courts as well as some other international courts (Amicus Curiae: European Court of Human Rights (ECtHR); Amicus Curiae: Inter-American Court of Human Rights (IACtHR); Amicus Curiae: African Court on Human and Peoples’ Rights (ACtHPR); International Courts and Tribunals, Amicus Curiae; Heri, 2018, paras 74–76). It is also true for the regional human rights commissions, which have not always made the point expressly, though Rules 104 and 105 ACommHPR’s 2020 Rules of Procedure do explicitly provide for amicus curiae submissions (see Piovesan and Cruz, 2020, paras 16–19; Weber, 2021, paras 15–19 and 21). The practice at the United Nations treaty bodies has been more diverse, though there is a strong trend towards the open submission of information (see Piovesan and Cruz, 2020, paras 22–25; Weber, 2021, paras 7–14 and 27–38).

24  State Parties have a duty to cooperate with international human rights courts in good faith, and human rights courts can, and often will, ask states to produce certain forms of evidence (see Kazazi, 1996, 120–21; Amerasignhe, 2005, 130–38; Heri, 2018, paras 51–53; Paúl, 2018, para 16; O’Boyle, 2018, paras 11–16). While the ability of quasi-judicial human rights bodies to demand compliance may be less stringent, there is no reason to understand states as less bound to engage in good faith with the procedures of those bodies whose jurisdiction to hear complaints they have accepted. Where states fail to engage in good faith in those procedures, for instance by failing to provide information to which they have access or failing to respond to a contention at all, that failure may be borne in mind in subsequent assessment of the evidence, including by reaching an assessment of the evidence in line with the contentions of the applicant (see, eg, Mursalov and ors v Azerbaijan, 2022, para 9.7).

25  In practice, human rights adjudication has frequently recognized the need not to place the entire burden of production and persuasion on claimants. Human rights bodies have expressed this point through two different but broadly parallel languages.

26  First, human rights adjudication has utilized inferences. ‘Inferences’ refers to the drawing of reasonable conclusions from indirect or circumstantial evidence. Human rights courts have stated that they will make ‘inferences’ where the situation suggests those inferences are appropriate (see Velasquez Rodriguez v Honduras, 1988, para 124; El-Masri v Macedonia, para 151; Merabishvili v Georgia, para 312).

27  Second, human rights courts and other human rights bodies have also referenced, and commentators have proposed, reversing the burden of proof in certain situations (see Roberts, 2021). Reversal of the burden of proof may be taken as a more formal means of achieving the outcome achieved through the drawing of inferences. As the below commentary makes clear, finding the appropriate balance in this area requires sensitivity to a range of considerations, including attention to relative access to information as well as underlying substantive policy goals.

28  One case that illustrates the close relationship between inferences and reversals of the burden of proof is the UNHRC decision in Q v Denmark. In that case, an Iraqi citizen who applied for naturalization in Denmark was denied a request for medical exemption from a language requirement. The claimant argued that his submissions ‘established a prima facie case through a number of strong, clear and concordant inferences and unrebutted presumptions’ that the denial of medical exemption had been unreasonable (Q v Denmark, 2015, para 5.5). For its part, the UNHRC found that the state had ‘failed to demonstrate that the refusal to grant the exemption was based on reasonable and objective grounds’ (Q v Denmark, para 7.5). Thus, the UNHRC in effect found for the claimant but did so through a finding the state had failed to make its case, rather than by explicitly agreeing with the claimant’s inferences—suggesting human rights bodies may prefer reverse burdens to inferences, insofar as reverse burdens allow human rights bodies to reach more cautious conclusions.

29  There are four categories of situations in which inferences and/or reversals of the evidentiary burden have a clear track record of being relied upon, the first in the context of the admissibility issue of exhaustion of local remedies (Local Remedies, Exhaustion of), the other three in the context of the merits.

30  Relative to admissibility, the IACommHR has indicated that, once claimants have advanced the case that they have exhausted domestic remedies or that an exception to the rule of exhaustion should apply, ‘the burden of proof with respect to the exhaustion of domestic remedies is on the state that alleges non-exhaustion’ (Parque São Lucas v Brazil, 2003, para 25; see also de Oliveira v Brazil, 1998, para 25; Ruiz Fuentes v Guatemala, 2008, para 57; Ecuador (on behalf of Franklin Guillermo Aisalla Molina) v Colombia, 2010, para 158). The ACommHPR (Recontre Africaine Pour la Defense des Droits de l’Homme v Zambia, 1996, para 12; Nubian Community in Kenya v Kenya, 2015, para 47; Geneviève v Cameroon, 2015, paras 83–85), the UNHRC (Kone v Senegal, 1994, para 5.3), and the CRC (AL v Spain, 2019, paras 5.3 and 11.3) have endorsed this principle as well.

31  Relative to the merits, there are three broad categories of situation in which reversals of the burden of proof have been deemed justifiable. First, there are situations in which one party has better access to the information in question. The need for a reversal of the burden here has been recognized by the ECtHR, which has held:

32  The reversal of the burden in this category of cases is justified because, as the IACtHR has held, ‘in proceedings to determine human rights violations the State cannot rely on the defense that the complainant has failed to present evidence when it cannot be obtained without the State’s cooperation. The State controls the means to verify acts occurring within its territory’ (Velasquez Rodriguez v Honduras, 1988, paras 135–36; see also Escher and ors v Brazil, 2009, para 127). The IACommHR has supported these points (see Abella v Argentina, 1997, para 196; Extrajudicial Executions and Forced Disappearances v Peru, 2001, para 184).

33  A similar point has been extensively recognized by other human rights bodies as well. The UNHRC, for instance, has observed that:

34  The CRC has similarly noted that ‘the burden of proof does not rest solely on the author of the communication, especially considering that the author and the State party do not always have equal access to the evidence and that frequently the State party alone has access to the relevant information’ (AM v Switzerland, para 10.5; see also COC v Spain, 2021, para 8.2; MT v Spain, 2019, para 13.4).

35  The CAT, meanwhile, has recognized a similar outcome via a different form of reasoning, observing:

In other words, in addition to taking into account the state’s enhanced ability to access certain information, human rights bodies may also take into account challenges in accessing, obtaining and bringing forward information faced by claimants.

36  Numerous fact patterns fall under the heading of situations in which one party—typically, the respondent state—has better access to information. The ECtHR has observed for instance that among such situations are those ‘concerning people in the custody of the authorities – in which the respondent State alone has access to information capable of corroborating or refuting the applicant’s allegations’ (Merabishvili v Georgia, para 313). The IACommHR has recognized this point as well. As the Commission has put it, ‘[t]he burden of proof lies with the State’ because ‘the State has exclusive control over information or evidence regarding the fate of the detained person’ (Morales Zegarra and ors v Peru, 1999, para 51; see also Extrajudicial Executions and Forced Disappearances v Peru, 2001, para 184).

37  The ECtHR’s jurisprudence suggests that situations of detention, in which the burden will fall on the state, should be understood to include situations ‘where, although it has not been proved that a person has been taken into custody by the authorities, it is possible to establish that he or she was officially summoned by the authorities, entered a place under their control and has not been seen since’, in regards to which the ECtHR has held ‘the onus is on the Government to provide a plausible and satisfactory explanation as to what happened … and to show that the person concerned was not detained’ (El-Masri v Macedonia, para 153; see also Khadzhialiyev and ors v Russian Federation, 2008, para 80).

38  The WGAD has also addressed this issue in one of its reports, where it observed that its evidential approach ‘is in line with the ruling of the International Court of Justice in Ahmadou Sadio Diallo (Republic of Guinea v Democratic Republic of the Congo)’ (WGAD, Report of the Working Group on Arbitrary Detention, 2011, para 68). That position, according to the WGAD, is one that recognizes the difficulties that may be faced in cases in which ‘it is alleged that a person has not been afforded, by a public authority, certain procedural guarantees to which he or she was entitled’; that ‘[a] public authority is generally able to demonstrate that it has followed the appropriate procedures and applied the guarantees required by law – if such was the case – by producing documentary evidence of the actions that were carried out’; and that therefore, after the claimant has established a prima facie case, ‘[i]n general the burden rests with the Government’ and ‘it is for the Government to produce the necessary proof’ (WGAD, Report of the Working Group on Arbitrary Detention, 2011, para 68). In the same context, the WGAD has indicated that ‘mere assertions that lawful procedures have been followed will not be sufficient to rebut the source’s allegations’ (WGAD, Report of the Working Group on Arbitrary Detention, 2011, para 68). The WGAD has complied with this well-reasoned approach in numerous subsequent cases, placing the burden of proof on the government after a prima facie case has been made.

39  In line with the above, differential access to information also suggests the burden should be on the state to disprove any prima facie case of death, torture, or cruel, inhuman, or degrading treatment or punishment while a claimant is in custody. This has been recognized by the IACtHR, which has held that:

40  For its part, in a case concerning allegations of mistreatment in custody, the IACommHR observed that ‘[a]lthough the burden of proof lay with the Mexican State in terms of the processing of this case by the Inter-American Commission, the Mexican State failed to meet its obligation to disprove the allegations made in a serious and well-founded manner’ (González Pérez and ors v Mexico, 2001, para 38). Moreover, as the IACommHR has noted elsewhere, in cases of detention:

41  The ACommHPR has similarly found the burden to shift to the respondent state in the context of ‘injur[y] in detention or while under the control of security forces’ (Egyptian Initiative for Personal Rights and Interights v Egypt, 2011, para 168). As the ACommHPR has put it elsewhere, ‘[w]hen a person is injured in detention or while under the control of security forces, there is a strong presumption that the person was subjected to torture or ill-treatment. The burden then shifts to the Respondent State to convince the Commission that the allegations of torture raised by the Complainants are unfounded’ (Okiring and Samson (represented by Human Rights Network and ISIS-WICCE) v Uganda, 2017, para 97).

42  As some of the language in the previous discussion suggests, reversal of the burden in such cases is linked not only to the state’s better access to information, but also both to the state’s obligation to ensure appropriate treatment of those detained, as well as to the state’s obligation to carry out investigations where serious violations have taken place (on states’ duty to investigate, see, eg, Heri, 2018, paras 10–18). The former point has been underscored by the IACommHR, which has observed that ‘[t]he burden of proof lies with the State … when the State holds a person in detention and under its exclusive control [because in those circumstances the state] becomes the guarantor of that person’s safety and rights’ (Morales Zegarra and ors v Peru, 1999, para 51; see also Extrajudicial Executions and Forced Disappearances v Peru, 2001, para 184). The latter point has been observed by the CERD, which has noted that in cases of serious allegations of violence and harm, ‘the onus [is] on the State party to initiate an effective criminal investigation, instead of giving the petitioners the burden of proof in civil proceedings’ (Dawas and Shava v Denmark, 2012, para 7.4). In general, reversing the burden in such cases makes sense, insofar as failure to place the burden on the state would undermine the state’s duty to investigate (Roberts, 2021, 1688–89).

43  Instances of differential access to information also include other situations in which relevant facts connected to ‘the events in issue lie wholly, or in large part, within the exclusive knowledge of the authorities’ (DH and ors v Czech Republic, 2007, para 179). These include situations involving the alleged use of excessive force by the authorities (see Soare and ors v Romania, 2011, para 139; Ataykaya v Turkey, 2014, para 46). The ACtHPR has pronounced a similar principle, by indicating that inferences will be drawn where ‘it is the Respondent state’s agencies which have the access to [the relevant] records’ (Gihana and ors v Rwanda, 2019, paras 84–85; see also Penessis v Tanzania, 2019, para 92). These include situations concerning official documents, relative to which the authorities have presumptive knowledge, as in Gihana. Where individuals have made a prima facie case to be nationals of a particular state, moreover, the ACtHPR has suggested the burden will then be on the respondent state to disprove that assertion, should they contest it (see Penessis v Tanzania, 2019, paras 96–97). The ACommHPR, meanwhile, has indicated that where the dispute concerns the extent of reasoning in an official judgment, which is in the hands of the authorities, ‘the burden of proof is on the Defendant to show that the allegations of the Complainants are unfounded by providing the Commission with the said judgement’ (Wetsh’okonda Koso and ors v Democratic Republic of Congo, 2008, para 90).

44  The CAT has been clear that situations of differential access to information also include situations in which applicants have difficulty accessing information needed to show there are substantial grounds for believing they would be in danger of being subjected to torture if expelled, returned or extradited. In its General Comment No 4 the CAT indicated that:

The CAT has subsequently followed this approach in its decisions (see, eg, X and Y v Switzerland, 2022, para 7.3).

45  Situations in which the authorities have better access to the relevant information also include situations in which court proceedings have been delayed. The IACtHR has indicated it will consider such delay unreasonable, failing the submission of explanatory information: ‘in certain cases, a prolonged delay may, in itself, constitute a violation of judicial guarantees. It is for the State to explain and prove why it has required more time than would be reasonable, in principle, to deliver final judgment in a specific case, in accordance with these criteria’ (Canese v Paraguay, 2004, para 142; Paúl, 2018, para 18). The UNHRC has come to a similar conclusion, noting that, once a prima facie case of delay has been made, it is ‘incumbent upon the State party to demonstrate that the particular circumstances of the case justified [the] prolonged’ process in question (Desmond Williams v Jamaica, 1997, para 9.4). The IACommHR has reached such a conclusion as well, observing relative to ‘pre-trial delay’ that:

46  Situations of differential access to information also include cases of discrimination, as alleged in DH (DH and ors v Czech Republic, 2007, para 3). In the case of Hoogendjik, the ECtHR indicated that, once claimants had established a prima facie case of discrimination, the burden would be on the state to disprove discrimination (Hoogendijk v Netherlands, 2005, 21–22; see also Heri, 2018, para 84; Cînta v Romania, 2020, para 79). For its part, the IACtHR, in recognition of ‘the difficulty for those who are the object of discrimination to prove racial prejudice’, has held that it ‘agrees with the ECtHR that, in certain cases of human rights violations motivated by discrimination, the burden of proof falls on the State, which controls the means to clarify incidents that took place on its territory’ (Dorzema et al v Dominican Republic, 2012, para 229). As Mackic has observed, the IACtHR’s statement would have captured the ECtHR’s finding better, and been more convincing overall, without the reference to ‘motive’ (2018, at 218–19). In terms of recognizing discrimination in practice, however, the European and IACtHRs have adopted similar positions. The ACommHPR has similarly observed that ‘in claims of discrimination, the burden of proof should be shared. The Complainant is required to present sufficient facts from which it can be presumed that discrimination may have occurred, after which the burden of proof shifts to the Respondent’ (Open Society Justice Initiative (on behalf of Noumeni) v Cameroon, 2019, para 189).

47  The CERD Committee has also addressed the issue of the burden of proof and discrimination, by calling for France to ‘ensure that the principle of reversal of the burden of proof is fully observed by: (a) enhancing the judicial procedures available to victims of racial discrimination by, inter alia, rigorously applying the principle of reversal of the burden of proof’ (Gabaroum v France, 2016, para 9). In practice, the CERD Committee here was calling on France to enforce the law it had in effect, not calling on a state without such a law to adopt, as a matter of obligation, rules requiring a reversal of the burden of proof. Nonetheless, human rights principles would seem to push in such a direction, and if such a principle applies on the national level, it should presumably apply on the international level as well (for some related considerations, see Kapotas, 2017).

48  The second category of situations relative to the merits in which a reversal of the burden has been deemed appropriate applies in situations in which the violation or violations in question form part of a widespread and/or systematic pattern of violations. The IACtHR has held that ‘[w]hen the existence of … a policy or practice has been shown, the disappearance of a particular individual may be proved through circumstantial or indirect evidence or by logical inference’ (Velasquez Rodriguez v Honduras, 1988, para 124). As one scholar has put it, where both ‘(a) the existence of a widespread violation of human rights; and (b) a link between this generalised practice and the particular case of an alleged victim’ can be shown, a reversal of the burden of proof is justified (Paul, 2015, 39).

49  The IACommHR has reasoned similarly. In the course of discussing reversal of the burden in the context of alleged enforced disappearances, the Commission observed that such a burden reversal was ‘even more important in view of the aforementioned government practice of causing disappearances’, before continuing to observe that ‘when there is proof of the existence of a policy of disappearances sponsored or tolerated by the Government, it is possible, using circumstantial or indirect evidence, or through relevant logical inference, to prove the disappearance of a specific individual when that would otherwise be impossible given the link between that disappearance and the overall policy’ (Morales Zegarra and ors v Peru, 1999, paras 48 and 50).

50  Insofar as reasoning along such lines has often corresponded to situations of enforced disappearance or the like, these cases might be understood as falling within the compass of the category of covert violations, discussed below. Merging the categories entirely would be a mistake, however, insofar as it would ignore the strong logic that pushes towards reversal of the burden in all cases of widespread or systematic violation, given the common-sense logic pushing towards the drawing of inferences in such cases and the accordance of such an approach with the policy goals of human rights systems (see Roberts, 2021, 1693–95).

51  The third category of situations relative to the merits in which human rights bodies have reversed the burden of proof is in the context of violations which are by nature clandestine, where proof is hard to come by. Such forms of rights violation include incommunicado detention, enforced disappearance and extraordinary rendition. The IACtHR, for instance, has held that ‘[c]ircumstantial or presumptive evidence is especially important in allegations of disappearances, because this type of repression is characterised by an attempt to suppress all information about the kidnapping or the whereabouts and fate of the victim’ (Velasquez Rodriguez v Honduras, 1988, para 131). The ACtHPR has held that:

52  Relative to extraordinary rendition, the ECtHR has held that given the ‘difficulties involved in gathering and producing evidence’ due to ‘the extreme secrecy surrounding … rendition operations’ it is particularly appropriate to rely on ‘circumstantial evidence, including a large amount of evidence obtained through … international inquiries, considerably redacted documents released by the CIA, other public sources and evidence from the experts and the witness’ in such contexts (Al-Nashiri v Poland, 2014, para 400).

53  Human rights bodies have reasoned similarly. The UNHRC, for instance, has reiterated its emphasis on the need for a distribution of the burden of proof due to informational issues arising in the context of enforced disappearance in particular (see Al Khazmi v Libya, 2013, para 8.2; Almegaryaf and Matar v Libya, 2014, paras 7.2 and 7.4; Al Daquel and ors v Libya, 2014, para 6.3). The ACommHPR has observed that ‘the context of the Victims in incommunicado detention and interrogation is such that available evidence is necessarily limited’ (Egyptian Initiative for Personal Rights and Interights v Egypt, 2011, para 142). The CAT may also be understood to have supported reversal in such cases, by observing:

54  The IACommHR has been most detailed on the subject of the appropriateness of inference and the need for a reversed burden in the context of enforced disappearances. To begin, the Commission has observed:

The Commission further clarified that this means that in such situations ‘it is not incumbent upon the petitioners to prove that the victims have disappeared, because it may be assumed, for lack of proof to the contrary, that the … State is responsible for the disappearance of any person it has detained’ (Morales Zegarra and ors v Peru, 1999, para 48). As the Commission explains, one justification for adoption of such an approach is the fact that a policy of enforced disappearance has as an aim and modality the concealment and destruction of evidence, depriving victims of the information they would otherwise need to advance their claims (Morales Zegarra and ors v Peru, 1999, para 49).

55  Scholars have also supported the drawing of inferences and/or reversal of the burden of proof in cases of enforced disappearances (Vermeulen, 2012, 213–52; Keller and Heri, 2014, 738 and 745–50; Roberts, 2021). The justification for reversal of the burden is particularly strong in such cases as, in addition to being justified due to the purposefully clandestine nature of the activities in question, they typically involve both differential access to information and situations of widespread and systematic patterns of violation as well.

56  Another factor human rights bodies might reasonably take into account in their consideration of evidence is the vulnerability and marginalization of human rights claimants in many cases, and the challenges posed by such conditions to claimants’ ability to develop their cases. The ECtHR has acknowledged such factors, by observing it would be ‘sensitive to any potential evidentiary difficulties encountered by a party’ (Merabishvili v Georgia, para 315).

57  Limitations faced by claimants may relate to their capacity, including for instance where claimants are young children. Vulnerability may relate to past experiences—for instance, the rights abuses claimants have suffered may leave them psychologically traumatized, without resources, and/or refugees or asylum seekers in a foreign state. Vulnerability may also relate to membership in a minority, vulnerable, marginalized, and/or discriminated against group. As the ECtHR has held, ‘the vulnerable position’ of particular communities ‘means that special consideration should be given to their needs … in reaching decisions in particular cases’ (DH and ors v Czech Republic, 2007, para 181). While this holding was in relation to national level measures, the same logic should apply at international human rights bodies.

58  Vulnerability may also relate to present and future circumstances, as those bringing human rights claims to the international level often face domestic circumstances that are inhospitable to their claims and may be at serious risk of reprisals and/or further rights violations, including due to making human rights claims and seeking the evidence to support those claims. These also constitute important factors for human rights bodies to bear in mind when adjudicating on issues of proof.

59  Certain rules adopted by human rights bodies demonstrate sensitivity to such issues. For instance, most of the United Nations human rights bodies have endorsed the San José Guidelines against Intimidation or Reprisals (2015). Human rights bodies have yet to clearly and extensively address such issues in the context of their approaches to proof, however.

60  As noted at the beginning, human rights bodies have individually been less than perfectly clear as to the approach to proof they employ. When the work of human rights bodies is examined collectively, however, the outlines of several clear and well-reasoned approaches are apparent. While perhaps not always applied consistently, human rights bodies have collectively developed a detailed, sensitive, and flexible framework that aims to capture the need for variations in the manner in which the burden of proof is distributed based on the types of human rights violation alleged.

Christopher M Roberts