Docket: IMM-15037-23
Citation: 2025 FC 1943
Ottawa, Ontario, December 9, 2025
PRESENT: Mr. Justice McHaffie
|
BETWEEN: |
|
SHETERA BROWN
KEITH DUVAL ELDON JR |
|
Applicants |
|
and |
|
THE MINISTER OF CITIZENSHIP AND IMMIGRATION |
|
Respondent |
JUDGMENT AND REASONS
I. Overview
[1] Shetera Brown and her son seek Canada’s protection from her ex-partner in the Bahamas, who physically abused and assaulted her while they were together and continued to threaten her after she ended their relationship in December 2020. The Refugee Appeal Division [RAD] of the Immigration and Refugee Board of Canada [IRB] rejected their claim, finding they had not rebutted the presumption that state authorities in the Bahamas could adequately protect them.
[2] On this application for judicial review, the applicants allege that the RAD’s analysis of state protection was unreasonable and ask the Court to quash the decision. In particular, they argue the RAD drew an unreasonable factual inference regarding the Bahamian police response to a report Ms. Brown filed in September 2020; undertook a selective examination of the objective evidence regarding protection from gender-based violence in the Bahamas; improperly relied on evidence of state efforts to combat gender-based violence rather than the effectiveness of those efforts; and inappropriately considered the likelihood of future risk in assessing state protection.
[3] Having reviewed the applicants’ arguments and the evidence in detail, I conclude the applicants have not met their onus to establish that the RAD’s decision was unreasonable. The RAD reasonably assessed the evidence of Ms. Brown’s interactions with the police and the objective evidence of conditions in the Bahamas, applied the correct legal framework, and appropriately considered whether the state could adequately protect Ms. Brown from the particular risks she faced. The applicants’ arguments, and the passages they cite from the evidence and the decision, do not show that the RAD fundamentally misapprehended or failed to consider the evidence, or that it applied the wrong approach to state protection.
[4] The application for judicial review will therefore be dismissed.
II. Issues and Standard of Review
[5] The parties agree that the RAD’s assessment of the availability of state protection is subject to judicial review on the reasonableness standard: Canada (Minister of Citizenship and Immigration) v Vavilov, 2019 SCC 65 at paras 16–17, 23–25; Burai v Canada (Citizenship and Immigration), 2020 FC 966 at para 17. The only issue on this application is therefore whether the RAD’s decision was reasonable.
[6] To answer this question, the Court reviews the RAD’s decision and reasons in the context of the evidence and the parties’ submissions to assess whether it is internally coherent and rational, and whether it is transparent, intelligible, and justified in light of the relevant factual and legal constraints: Vavilov at paras 15, 82–101, 125–128. Reasonableness is a deferential standard, but it is not a rubber-stamping process; it remains a robust form of review: Vavilov at para 13. To establish that an administrative decision is unreasonable, however, an applicant cannot simply point to a superficial flaw or a “minor misstep.”
They must show that the shortcomings in the decision are so sufficiently serious and central that it does not exhibit the requisite degree of justification, intelligibility, and transparency: Vavilov at para 100.
[7] The goal of reasonableness review is not for the Court to re-decide the issue based on its own assessment and appreciation of the evidence: Vavilov at paras 83, 125–126. Rather, the Court’s role is limited to assessing whether the decision maker has made a fundamental error in logic or justification, in its appreciation of the evidence, or in its application of the relevant law: Vavilov at paras 96, 99–102, 105–107.
III. Analysis
A. The RAD did not engage in unreasonable speculation
[8] The applicants argue that the RAD relied on mere speculation in its assessment of an interaction Ms. Brown had with the Royal Bahamas Police Force in September 2020. For the following reasons, I conclude that the applicants have not shown that the RAD based its decision on unreasonable or improper speculation.
(1) Factual context: the basis of claim and reports to the police
[9] The applicants’ refugee claim is based on fear of Ms. Brown’s former partner and the father of her son, identified by the RAD as “KE.”
Between October 2010 and December 2020, KE lived with Ms. Brown and their son at Ms. Brown’s apartment. Over time, the relationship deteriorated, and KE became more confrontational and then violent. KE assaulted Ms. Brown in 2018. In 2020, he began physically abusing her on a regular basis.
[10] On September 2, 2020, after KE assaulted Ms. Brown, she reported the incident to the Royal Bahamas Police Force, providing a written statement and KE’s name, contact number, and location. The police told her they were going to call KE and that they would call her back once they had anything. However, they never called her back. Ms. Brown did not follow up further with the police, because she was waiting for their call. Ms. Brown’s impression was that the police were unsympathetic, and that their disregard may have been motivated by racism, since KE is white. She also believed that the police did not speak with KE after this report, and it appears he was neither arrested nor charged.
[11] In December 2020, Ms. Brown travelled to Florida for a few days to shop. On her return, KE beat her, claiming she had intentionally not activated her data roaming. This led Ms. Brown to end their relationship and kick him out of her apartment at the end of 2020. Ms. Brown accepted a job on a different island and her son moved in with Ms. Brown’s mother in Nassau. KE continued to call and text Ms. Brown with threats, including threats to kill her if he found out she was seeing someone else and threats to come to the island where she was working.
[12] In July 2021, KE threatened to take their son away from Nassau and off the island. Ms. Brown contacted the police again. They informed her that KE could not take their son without her consent and that if he did, “they would have locked him up.”
The police contacted KE and warned him he would be locked up if he took their son. KE did not take their son, but he was very angry with Ms. Brown, threatening her on a call and boasting about his connections with the police.
[13] In early 2022, Ms. Brown was planning a trip home to Nassau to see her son. However, KE found out about the planned trip and sent her texts threatening to kill her. She decided to flee the Bahamas and did so with her son shortly thereafter.
(2) The RAD’s consideration of Ms. Brown’s interactions with police
[14] In assessing the availability of state protection, the RAD considered Ms. Brown’s interactions with the police in September 2020 and July 2021. With respect to the 2020 report, the RAD found Ms. Brown’s assessment that the police had not spoken with KE, and that they might have been unhelpful based on their racism, to be speculative, since she had no information about any conversation the police might have had with KE.
[15] The RAD did not agree that the lack of a follow up call from the police meant that they had taken no action at all. Rather, the RAD found that it “suggests the police had nothing conclusive to communicate.”
The RAD noted that Ms. Brown did not follow up with the police or make any inquiry as to the status of any pending investigation. The RAD recognized that Ms. Brown genuinely believed the police took no action, but found there was no reliable and convincing evidence of “their action, or inaction, at that time.”
(3) The RAD’s findings on this issue were reasonable
[16] The applicants argue that the RAD’s inference that the police did not call back in 2020 was because they had nothing conclusive to communicate was speculative and unreasonable. They question how the police could have nothing to communicate when they had KE’s name and contact information for their investigation. They also point to KE’s angry reaction when the police did contact him in 2021, arguing that the absence of such a reaction in 2020 suggests that the police did not contact him then. The applicants argue that the inference that they draw from these known facts—that the police took no action at all—should have been accepted as more consistent with the evidence.
[17] It is important to underscore two principles from the case law with respect to inferences and speculation. The first is that there is no “bright line”
between the two, such that distinguishing between permissible reasonable inferences and impermissible speculation or conjecture is often very difficult: Ukleina v Canada (Citizenship and Immigration), 2009 FC 1292 at para 14, citing Jones v Great Western Railway Co (1930), 47 TLR 38 (HL) at p 45; KK v Canada (Citizenship and Immigration), 2014 FC 78 at paras 60–61, citing R v Munoz, 2006 CanLII 3269 (ON SC) at paras 23–26, 28–31; R v Villaroman, 2016 SCC 33 at paras 38–43. The second is that the task of drawing inferences from the evidence is the job of the trier of fact—here the RAD—rather than a reviewing court: Vavilov at para 125; Asowata v Canada (Citizenship and Immigration), 2023 FC 462 at para 13; Iqbal v Canada (Immigration, Refugees and Citizenship), 2025 FC 759 at para 15; Contact Lens King Inc v Canada, 2022 FCA 154 at para 83. While a court on judicial review can and must assess whether an inference is reasonable or speculative, this role is to be performed in the context of the deference owed to the RAD, and the general rule that a reviewing court is not to reassess the evidence and draw its own conclusions: Ukleina at paras 7–8, 14; Vavilov at paras 15, 83, 125.
[18] In the present case, the known facts established by Ms. Brown’s testimony were that she reported the assault to the police; that the police told her they would call KE and that they would contact her once they had anything (or “if something [came] out of it”
); and that the police did not contact her further. The conclusion that Ms. Brown drew from this, and that the applicants urge on the Court, was that the police did nothing at all in response to her report. The RAD was not prepared to draw this inference.
[19] The RAD noted that the lack of a call to Ms. Brown “suggests”
the police had nothing conclusive to communicate to her, and later in its reasons stated that the police “initiat[ed] some manner of investigation.”
However, the RAD’s primary conclusion was simply that the evidence did not support Ms. Brown’s belief that the police took no action at all. In this regard, the RAD noted that the police’s “action, or inaction”
in 2020 was not supported by reliable and convincing evidence, indicating that it was not making a positive finding either way.
[20] Thus, the only inference the RAD appears to have drawn was that the police initiated some manner of investigation, but not one that resulted in a call to Ms. Brown. Given the deference owed to the RAD’s assessment of the evidence, I conclude that this limited inference was one that was open to the RAD and was not simply speculation based on no facts. While other inferences might also have been available, including that put forward by Ms. Brown, this does not mean that the inference the RAD drew was unreasonable or amounted to mere conjecture.
[21] Even if one reads the RAD’s decision as making a positive finding that the police did not contact Ms. Brown because they had nothing conclusive to communicate, this was not itself an unreasonable inference. The police had told Ms. Brown that they would contact her when they had information. The fact that they did not contact her could lead to the inference that they did nothing at all, or to the inference that they did something, but not enough to result in information to communicate. In my view, it was open to the RAD to draw the latter inference from the facts rather than the former, without engaging in speculation.
[22] With respect to the question of the police’s asserted failure to contact KE, the RAD’s reasons, read as a whole, do not indicate it drew a factual inference that the police did in fact contact KE. Again, the RAD’s observation that the police’s “action, or inaction”
was not supported by reliable and convincing evidence indicates that it did not make such a finding.
[23] I make two other notes in respect of the applicants’ argument that an inference should be drawn from the lack of any apparent angry reaction from KE in 2020. The first is that this issue was not raised with the RAD. In their memorandum of argument before the RAD, the applicants referred to Ms. Brown’s interactions with police in 2020 and 2021. However, they focused on Ms. Brown’s beliefs regarding the adequacy of state protection and the difference between police protection for child abduction (the 2021 interaction) and for domestic violence (the 2020 interaction). They did not argue to the RAD, or to the Refugee Protection Division [RPD], that KE’s lack of reaction was proof of a lack of contact or other lack of action. The RAD cannot be faulted for not having addressed this argument or line of reasoning in its decision.
[24] The second note is that there is an inconsistency in the applicants’ arguments regarding KE’s conduct. They argue that since KE reacted angrily to the 2021 police interaction, one can assume that he would have reacted similarly to police contact in 2020, and that the lack of such reaction indicates that the police did not contact him. At the same time, however, they argue that a similar conclusion that the RAD drew—that KE’s compliance with the police’s warning in 2021 meant it was unlikely he would disregard a formal restraining order—amounted to unreasonable speculation as to the “logic of the agent of persecution,”
which this Court has criticized: Senadheerage v Canada (Citizenship and Immigration), 2020 FC 968 at para 19. I will address this matter further below, but note for the present that an applicant cannot argue that their agent of persecution will act consistently when it suits their narrative, but accuse the RAD of improper speculation when it draws a similar inference of consistent behaviour.
[25] The applicants reasonably point out that regardless of any investigation or contact that may have occurred, the September 2020 report to the police resulted in no charges, protection, or follow-up with Ms. Brown. I agree that regardless of whether the police conducted some amount of investigation, contacted KE, or simply took the report and did nothing further, the police response was not a sufficient response to a complaint of intimate partner violence.
[26] However, the question becomes how that incident, together with the remainder of the evidence relevant to state protection, is considered in assessing whether Bahamian authorities could adequately protect Ms. Brown in the future. That issue, which is a different one from whether the RAD made an unreasonable or speculative inference from the evidence, is addressed further below.
[27] I therefore conclude that the applicants have not established that the RAD entered into unreasonable speculation regarding the September 2020 report to the police.
B. The RAD reasonably assessed the evidence of state protection
[28] The applicants argue the RAD erred in two ways in assessing the country condition evidence regarding the availability of state protection: (1) by engaging in a selective review of the evidence and discounting information contrary to its conclusions; and (2) by relying on evidence of efforts on the part of the Bahamas rather than operational adequacy.
[29] It is worth noting before turning to these arguments that the applicants do not argue that the RAD misstated the law with respect to state protection. The RAD correctly noted at the outset of its discussion that international refugee protection is available as “surrogate or substitute protection”
when the country of nationality is unable or unwilling to provide that protection: Canada (Attorney General) v Ward, 1993 CanLII 105, [1993] 2 S.C.R. 689 at pp 709, 711–712, 716–717. Absent a situation of “complete breakdown of state apparatus,”
it is presumed that the state is capable of protecting a claimant: Ward at p 725. This presumption may be rebutted where there is relevant, reliable, and convincing evidence that demonstrates, on a balance of probabilities, that state protection is inadequate: Flores Carrillo v Canada (Minister of Citizenship and Immigration), 2008 FCA 94 at paras 1, 17–20, 24–30, 38.
[30] The RAD also referred to this Court’s decision in Gonzalez Torres for the proposition that a contextual approach should be taken to assessing whether a claimant has rebutted the presumption of state protection: Gonzalez Torres v Canada (Citizenship and Immigration), 2010 FC 234 at para 37, citing Garcia v Canada (Minister of Citizenship and Immigration), 2007 FC 79 and Avila Ortega v Canada (Citizenship and Immigration), 2009 FC 1057. The RAD set out the factors that Justice Zinn enumerated as relevant in Gonzalez Torres, namely (1) the nature of the human rights violation; (2) the profile of the alleged human rights abuser; (3) the efforts that the victim took to seek protection from authorities; (4) the response of the authorities to requests for their assistance; and (5) the available documentary evidence: Gonzalez Torres at para 37. Again, the applicants do not contend that the RAD misstated the law.
[31] Rather, the applicants’ two arguments pertain solely to the RAD’s review of the evidence. For the following reasons, I conclude that the RAD’s assessment of the availability of adequate state protection evidence did not fundamentally misapprehend the evidence, fail to account for relevant evidence, or otherwise fail to conform to the applicable factual and legal constraints: Vavilov at paras 105, 112, 125–126
(1) The RAD did not engage in a selective reading of the evidence
[32] The applicants argue the RAD’s assessment of the objective country condition evidence was selective, relying too heavily on statements that supported its conclusion while failing to grapple with the contrary evidence. They refer to and reproduce a large number of passages from the country condition evidence, both from the National Documentation Package [NDP] for the Bahamas published by the IRB and from other sources, that they had cited in their submissions to the RAD. They assert that this evidence indicates that adequate protection was not available to victims of domestic violence and that the RAD failed to consider it.
[33] For the reasons below, the applicants’ arguments and evidentiary references do not persuade me that the RAD unreasonably assessed the objective country condition evidence. Before turning to this issue, however, I will make a preliminary observation regarding the applicants’ references to documents in the NDP, not all of which were in the record before this Court. I make this observation largely as a practice point, as it is not determinative in this case.
(a) Preliminary observation: documents from the NDP not in the record
[34] A number of the passages cited by the applicants come from sources in the NDP that are not found either in the certified tribunal record [CTR] produced by the RAD or elsewhere in the record of this application for judicial review. The applicants have reproduced passages from these sources in their memorandum of argument, but it is difficult for the Court to fully assess this evidence without the context given by the documents themselves. It is therefore worth underscoring some relevant principles regarding the NDP and its status as part of the record.
[35] Documents contained in the NDP are considered to be available to the parties and the IRB and are part of the record before the IRB: Benavides Quispe v Canada (Citizenship and Immigration), 2021 FC 791 at para 22; Guzman v Canada (Citizenship and Immigration), 2024 FC 433 at paras 25–27; Siddique v Canada (Citizenship and Immigration), 2022 FC 964 at para 24; Magonza v Canada (Citizenship and Immigration), 2019 FC 14 at paras 77–79. Thus, for example, a document from the NDP is not considered a “new”
document before the RAD even if it was not referred to by or before the RPD: Benavides Quispe at paras 21–22.
[36] However, documents are not automatically part of the record before this Court simply by virtue of being part of the NDP, even if they have been referred to by a party, by the RPD, or by the RAD. The record on an application for judicial review in the citizenship, immigration, and refugee areas consists of (i) the applicant’s record and the respondent’s materials filed at the leave stage; (ii) the CTR produced by the tribunal pursuant to either a production order or an order granting leave; and (iii) any further materials filed by the parties pursuant to an order granting an application for leave: Federal Courts Citizenship, Immigration and Refugee Protection Rules, SOR/93-22 [IMM Rules], Rules 10(2), 11(a), 14(2), 15(1)(b)–(c), 17; López v Canada (Citizenship and Immigration), 2019 FC 349 at para 6; Amended Consolidated Practice Guidelines for Citizenship, Immigration, and Refugee Protection Proceedings, paras 33, 41. If a document, including a document from the NDP, is not in one of these places, it is not before the Court for consideration on an application for judicial review.
[37] In this regard, Rule 17(b) of the IMM Rules provides that the CTR prepared by the IRB is to include “all relevant documents”
in the possession of the tribunal: Canadian Association of Refugee Lawyers v Canada (Citizenship and Immigration), 2024 FC 128 at para 14. This does not include the entire NDP, as not every document in the NDP will be relevant. As a matter of practice, the CTR will typically include a list of the documents in the NDP, together with full copies of some documents.
[38] Should a party wish to rely on or refer to a document from the relevant NDP that is not in the CTR, or in materials filed at the leave stage, they must ensure it is before the Court by attaching it to a further affidavit filed in accordance with the order granting leave. Alternatively, a motion might be brought seeking a more complete CTR from the tribunal, although in most cases, it will be more efficient to simply introduce the document by affidavit.
[39] In accordance with what appears to be the RAD’s current practice in the preparation of CTRs, the CTR in this matter contains a section entitled “Items from National Documentation Package(s) cited in the RAD/RPD reasons.”
The title page of this section contains the following indication:
*URLs for documents are provided in the attached lists. Documents have been included where the URLs are invalid.
[40] I make no comment on whether the RAD’s approach of reproducing in the CTR only copies of documents where hyperlinks are invalid complies with Rule 17(b) of the IMM Rules.
[41] Importantly, however, a hyperlink—whether in the CTR or in a party’s materials—does not make the linked document part of the record of this Court on judicial review. Leaving aside obvious concerns about broken links and updated or modified documents, the Court is not responsible for, and is arguably precluded from, seeking out copies of documents from sources beyond the record, including from the NDP issued by the IRB. It is essential for accuracy, efficiency, and fairness that the Court and the parties know precisely the scope and extent of the record on the application for judicial review, including the contents of the documents being relied on. Unless the Court orders otherwise, parties should not treat a document that is merely listed and/or hyperlinked but not reproduced as being part of the record on an application for judicial review.
[42] In the present case, the NDP for the Bahamas was updated between the RPD’s decision and the RAD’s decision. The RPD referred to the May 31, 2022, version of the NDP. The NDP was updated on September 29, 2023, shortly after the applicants perfected their appeal to the RAD. The RAD notified the applicants of this update and invited them to make additional submissions on the updated NDP. They did not do so. The result of the update to the NDP and the RAD’s approach to preparing the CTR is that the CTR before this Court includes two lists of documents with hyperlinks and copies of three documents from the NDP, namely item 5.8 from the May 31, 2022, version (which was removed from the September 29, 2023, version), and items 2.1 and 5.3 from the September 29, 2023, version of the NDP (item 2.1 is an updated version of a US Department of State report found in the earlier NDP; item 5.3 is common to both versions of the NDP).
[43] The applicants’ references to the country condition evidence contain one sentence from item 2.1 and a longer passage from item 5.3. They also include passages ranging from one sentence to multiple paragraphs from six other items in the NDP that are not reproduced in the CTR, as well as lengthy passages from other country condition documents not from the NDP (these latter documents are reproduced in the CTR).
[44] For the reasons below, I conclude that even accepting the passages from the six unreproduced NDP documents on their face, without the context of the full document, and taking those passages together with the evidence reproduced in the CTR, the applicants have not established that the RAD’s assessment of the country condition evidence was unreasonable. The concern referred to above regarding the absence of these six documents from the record is therefore not determinative in this matter.
(b) The RAD’s assessment of the country condition evidence
[45] In addition to considering Ms. Brown’s own interactions with police in 2020 and 2021, the RAD considered the country condition evidence regarding state protection in the Bahamas. It found that despite Ms. Brown’s beliefs, the country evidence did not support a finding that state protection is operationally inadequate.
[46] The RAD referred to the RPD’s acknowledgement of deep-rooted patriarchal stereotypes in the Bahamas, and the treatment of violence against women and girls as a private matter. It recognized the applicants’ reference to worsening gender-based violence during the COVID-19 pandemic, but also noted that the same document the applicants relied on (i.e., the Department of State report at item 2.1) stated that the government “generally enforced”
the law on gender-based violence despite reports of police reluctance to intervene. The RAD went on to consider a number of other documents that referred to the legal obligation to protect women and girls from domestic violence and the availability of protection orders, noting that they are “enforceable by the police”
but that the police do not always inform women that they can apply for such an order.
[47] The RAD found that the evidence suggested that adequate state protection was available in the Bahamas, as the laws are “generally”
enforced, and that the reference to police “not always”
informing women of the right to apply for protection orders indicated that the norm was enforcement. The RAD recognized that state protection was not “without fault,”
and noted the applicants’ citation of statistics showing very high rates of violent crime and violence against women. However, the RAD cited Federal Court of Appeal jurisprudence holding that state protection need not be perfect to be operationally adequate: Zalzali v Canada (Minister of Employment and Immigration), 1991 CanLII 13557, [1991] 3 FC 605 (CA) at p 615; Canada (Minister of Employment and Immigration) v Villafranca, 1992 CanLII 8569 (FCA); see also Poczkodi v Canada (Immigration, Refugees and Citizenship), 2017 FC 956 at para 37.
[48] The RAD noted that the Bahamas has made and continues to make serious efforts against gender-based violence. It also referred to the police having initiated some manner of investigation in September 2020, and having dissuaded KE from taking action in 2021, thereby appearing responsive to Ms. Brown’s complaints despite her belief that they were dismissive. The RAD therefore did not accept that state protection was inadequate “because the only evidence suggesting this”
was Ms. Brown’s subjective belief.
(c) The RAD’s analysis was not unreasonably selective
[49] This Court has frequently recognized that a decision maker cannot limit their assessment of country condition evidence to a selective reliance on elements supporting their conclusion, while unjustifiably ignoring relevant and material contrary evidence: Whyte v Canada (Citizenship and Immigration), 2023 FC 1420 at paras 3, 26; Magonza at paras 84–93; B381 v Canada (Citizenship and Immigration), 2014 FC 608 at para 50; Cepeda-Gutierrez v Canada (Minister of Citizenship and Immigration), 1998 CanLII 8667 (FC) at paras 14–17; Vavilov at paras 125–126. The applicants argue that the RAD committed this error in its analysis of state protection.
[50] The applicants assert that the country condition evidence they cited to the RAD indicates that adequate protection is not available to survivors of domestic violence in the Bahamas. They contend that the RAD ignored the import of this evidence and failed to engage with evidence contrary to its conclusion. In particular, the applicants argue that the RAD’s decision “selectively focuses”
on a single line from the US Department of State report (item 2.1) about laws on domestic violence being generally enforced, while ignoring evidence to the contrary. They also assert that the RAD did not mention any of the evidence that “directly contradicts”
its assessment of the adequacy of state protection.
[51] Having reviewed the evidence cited by the applicants in detail and the RAD’s reasons, I cannot accept the applicants’ submissions. Contrary to the applicants’ arguments, the RAD did not simply focus or rely on a single line from a single document from the NDP while ignoring any contrary evidence. Rather, the RAD expressly identified concerns expressed in the documents cited by the applicants about societal stereotypes and the evidence of wide-spread gender-based violence in the Bahamas. It also referred to the legislative context, concerns about police reluctance, and the availability of protection orders, all of which were the subject of a number of the passages cited by the applicants. The RAD quoted from and gave citations to multiple documents from the NDP, in addition to referring to Ms. Brown’s own experiences.
[52] The applicants concede that the RAD quoted parts of the evidence and acknowledged evidence of high rates of gender-based violence. They nonetheless claim that the RAD’s reasons do not “engage with”
this evidence and “ignore its import.”
In essence, the applicants appear to be arguing that the RAD’s decision was unreasonable because it did not accept the applicants’ contention that the evidence showed a lack of adequate state protection. However, mere disagreement with the RAD’s conclusions does not show a decision to be unreasonable. The RAD considered both evidence related to the effectiveness of state protection in the Bahamas and evidence regarding its limitations, without ignoring or minimizing the contrary evidence. While noting that the evidence showed that state protection was not perfect or without fault, the RAD was satisfied that it met the standard of adequate state protection. This analysis was reasonable based on the evidence before the RAD.
[53] The applicants also point to the RAD’s statement that “the only evidence”
suggesting inadequate state protection was Ms. Brown’s subjective belief. They argue that this shows the RAD ignored the various aspects of the evidence they cited showing deficiencies in state protection. I cannot agree. The RAD’s reference to “the only evidence”
cannot be isolated from the context in which it appears in the RAD’s decision, which follows four paragraphs of discussion of the evidence regarding state protection, both positive and negative, as discussed above. Reading this passage in context, it is clear that the RAD’s conclusion was that the objective evidence, on balance, did not demonstrate that state protection was inadequate, and that therefore the only evidence of inadequate state protection was Ms. Brown’s belief.
[54] While a decision maker cannot conduct a one-sided review of country condition evidence, they are not obliged to refer to every piece of evidence that an applicant may cite or that may bear on an issue: Cepeda-Guttierez at para 16; Ciuron v Canada (Citizenship and Immigration), 2024 FC 97 at para 29; Del Pilar Platin Vargas v Canada (Citizenship and Immigration), 2014 FC 484 at para 25. I conclude that the RAD reasonably engaged with the evidence before it and neither fundamentally misapprehended nor failed to account for material contradictory evidence in its assessment of state protection.
(2) The RAD did not unreasonably rely on efforts in place of operational adequacy
[55] As the applicants correctly point out, the relevant question in assessing state protection is whether the state can actually protect its citizens as a matter of “empirical reality”
: Kumati v Canada (Citizenship and Immigration), 2012 FC 1519 at paras 27–28. As a result, evidence that a state has made “efforts”
to protect its citizens cannot be used as a substitute for evidence of effective adequate protection at an operational level: Kumati at para 39; Magonza at para 72; Whyte at para 21.
[56] The applicants point to the RAD’s reference to the Bahamas continuing to make “serious efforts against gender-based violence,”
and contends that the RAD fell into the error described in the foregoing cases by adopting a “serious efforts”
test for state protection rather than one of operational adequacy.
[57] I am not persuaded. The jurisprudence does not suggest that simply referring to a state’s efforts to protect its citizens or combat crime renders a decision unreasonable. Rather, the problem arises where there is a “focus on state efforts without an assessment of the effectiveness of those efforts”
: Whyte at para 21; Munzembo v Canada (Citizenship and Immigration), 2021 FC 1158 at para 27; Mata v Canada (Immigration, Refugees and Citizenship), 2017 FC 1007 at para 13.
[58] In the present case, the RAD referred several times to operational adequacy, showing its awareness that this was the relevant question assessing state protection. Its analysis then addressed operational issues, while also addressing efforts and improvements. Notably, as the applicants themselves underscore, the RAD relied on evidence indicating that the government generally enforced the law on gender-based violence, a clear reference to actual operational enforcement and not simply efforts. It also addressed the actual issuance and enforcement of protection orders, while recognizing operational limitations in the police’s failure to inform women about the availability of such orders.
[59] The RAD’s assessment of operationally adequate state protection also considered Ms. Brown’s own experiences with police, consistent with a contextual analysis that considers the efforts a claimant took to seek protection and the response of authorities to those requests for assistance: Gonzalez Torres at para 37. The RAD considered both the incident in 2020, which resulted in no charges against KE and no further contact from police, and the later event in 2021, which resulted in the police warning KE not to take their son away. The RAD found that this evidence confirmed that the police were not unwilling to act against KE. In this regard, I note that even accepting that the police response in 2020 was inadequate, this Court has found that a single negative interaction with police may be insufficient to rebut the presumption of state protection: Lesi v Canada (Citizenship and Immigration), 2016 FC 441 at para 47, citing Ruszo v Canada (Minister of Citizenship and Immigration), 2013 FC 1004 at para 51; Navarro Serrate v Canada (Citizenship and Immigration), 2025 FC 1293 at para 19.
[60] Given that the RAD directly addressed both general evidence regarding operational effectiveness and Ms. Brown’s individual experiences in seeking police assistance, I am not convinced that the RAD unreasonably focused its analysis on state efforts without undertaking the necessary assessment of the operational adequacy of state protection in the Bahamas.
C. The RAD reasonably considered Ms. Brown’s risk
[61] The applicants argue that the RAD erred by factoring into its state protection analysis an assessment of the particular future risk that Ms. Brown might face from KE. In particular, they point to (a) the RAD’s observations about KE’s past conduct and current unknown whereabouts, and (b) the RAD’s assessment of the likelihood of KE respecting a restraining order.
[62] On the former, the applicants point to the RAD’s statements about KE not having acted on his threats in the past, and the applicants’ inability to locate KE for the purposes of an Ontario family law proceeding. They argue the RAD effectively concluded that since KE had not acted on his threats in the past, Ms. Brown was not at risk, and then considered this in assessing the adequacy of state protection. They assert that the RAD, having found that the determinative issue was state protection, had to limit their analysis to the question of whether state protection would be available if Ms. Brown required it. By analogy to the analysis of an internal flight alternative, they argue that the question of forward-looking risk should not be conflated with that of state protection: Senadheerage at paras 49–50.
[63] I cannot agree. There is no indication that the RAD factored the likelihood of KE reappearing and seeking to harm Ms. Brown into their assessment of whether the police could and would protect her if she did. As the applicants themselves admit, the RAD did not “flesh out the import of the statements”
referred to above. Rather, their argument is founded on the applicants “wondering whether the panel factored in a lack of future risk into its state protection analysis”
[emphasis added]. An assessment of reasonableness must be driven by the reasons provided by the RAD, not speculation as to whether the RAD might have factored something into their analysis that their reasoning does not indicate.
[64] Again, the RAD’s reasons show it was aware that an assessment of adequate state protection requires a contextual analysis that considers the nature of the persecution and the profile of the agent of persecution, since state protection “cannot be determined in a vacuum”
: Gonzalez Torres at para 37; Avila Ortega at para 24. The nature of the protection required, and the adequacy of the protection available, may well be different depending on the nature and seriousness of the risk a claimant is facing and the profile of their agent of persecution. It was therefore not unreasonable for the RAD to consider Ms. Brown’s particular circumstances, including the nature of the harm she faced and the profile of KE, in assessing whether the protection available from the Bahamas would be adequate. Indeed, it may have been unreasonable for the RAD not to have conducted a personalized inquiry as to whether state protection was available in Ms. Brown’s particular circumstances: Matthias v Canada (Citizenship and Immigration), 2023 FC 619 at paras 26–27, citing Asher Belle v Canada (Citizenship and Immigration), 2012 FC 1181 at para 20, and Da Souza v Canada (Citizenship and Immigration), 2010 FC 1279 at para 6.
[65] This brings me to the applicants’ final argument, namely that the RAD unreasonably speculated that KE would comply with a protection or restraining order. The applicants contend that this amounts to speculation as to both the police issuing a restraining order, and KE respecting it.
[66] I disagree, for two reasons. First, the RAD’s consideration of a restraining order was not speculation, but was based on the country condition evidence that such orders can be issued promptly and are enforceable by police. Second, it was reasonable for the RAD to consider whether KE was likely to respect such an order, as this was material to a personalized assessment of the availability of adequate state protection. Consideration of the past conduct of an agent of persecution, and what that past conduct indicates about their likely future conduct, is not inherently unreasonable speculation about the “logic of the agent of persecution”
or their “motives, means and future intentions,”
provided that it is based in the evidence: Mirzayev v Canada (Citizenship and Immigration), 2024 FC 449 at paras 7–8; Engenlbers v Canada (Citizenship and Immigration), 2022 FC 1545 at para 14; Soos v Canada (Citizenship and Immigration), 2019 FC 455 at paras 12–16. Indeed, Ms. Brown’s own claim for refugee protection—like many such claims, if not most—is premised on the presumption that KE’s past conduct is a key indicator of his future conduct, and thus of her forward-looking risk.
[67] It is also worth noting that the RAD’s conclusion on this issue came in response to Ms. Brown’s assertion that KE would not respect a restraining order. The RAD noted that it seemed unlikely that he would respect a verbal caution from the police out of fear of jail time and simultaneously disregard a restraining order facing similar consequences. The RAD noted that Ms. Brown’s assertion as to what KE might do when facing a restraining order was speculative, whereas his actual actions when cautioned by police was to comply. This is a reasonable assessment based on inferences from the evidence.
IV. Conclusion
[68] The applicants have not established that the RAD’s conclusion that they had not rebutted the presumption of operationally adequate state protection in the Bahamas was unreasonable. As the existence of adequate state protection in a country of nationality means there is no need for surrogate protection in Canada, the RAD reasonably concluded that the applicants are not Convention refugees or persons in need of protection. The application for judicial review must therefore be dismissed.
[69] Neither party asked that I certify a serious question of general importance. I agree that no question meeting the requirements for certification arises in the matter.
Email this Content