NOTE: THE CONFIDENTIALITY OF THE NAME OR IDENTIFYING

PARTICULARS OF THE APPELLANT AND OF HIS CLAIM OR STATUS

MUST BE MAINTAINED PURSUANT TO S 151 OF THE IMMIGRATION
ACT 2009. SEE
[Link]

IN THE HIGH COURT OF NEW ZEALAND

AUCKLAND REGISTRY

I TE KŌTI MATUA O
AOTEAROA TĀMAKI
MAKAURAU ROHE
CIV-2023-404-2192
[2024] NZHC 1999

UNDER the Immigration Act 2009

IN THE MATTER of an appeal against a decision of the

Immigration and Protection Tribunal of
New Zealand under s 245 of the
Immigration Act 2009

BETWEEN EF
Appellant

AND REFUGEE AND PROTECTION

OFFICER
Respondent

Hearing: 10 April 2024 (further submissions received 03 May

2024)
Appearances: S R G Judd, D A Manning and S H Lamain for Appellant
M Deligiannis and S M Perera for Respondent

Judgment: 19 July 2024

JUDGMENT OF WILKINSON-SMITH J

This judgment was delivered by me on 19/07/2024 at 3pm.

Pursuant to Rule 11.5 of the High Court Rules.

…………………………
Registrar/Deputy Registrar

Solicitors/Counsel:
M Deligiannis, Crown Law Office, Wellington
S M Perera, Crown Law Office, Wellington
S R G Judd, Barrister, Auckland
D A Manning, Barrister, Auckland
S H Lamain, Barrister, Auckland

EF v REFUGEE AND PROTECTION OFFICER [2024] NZHC 1999 [19 July 2024]
 Table of Contents Para No
Introduction [1]
Background [8]
The first Tribunal decision [14]
The second Tribunal decision [32]
The issues [54]
Discussion [55]
Result [94]

Introduction

[1] EF is a citizen of [redacted] and a New Zealand resident. He was served with
a deportation liability notice following conviction for his role in the importation of
drugs into New Zealand. His claims for refugee and protected person status were
rejected by a Refugee and Protection Officer and his appeals were dismissed by the
Immigration and Protection Tribunal.

[2] Subsequently, the Court of Appeal granted leave to appeal on the following
question of law:1

Did the tribunal err in its approach to risk assessment and, as a result, did it
improperly exclude material information from consideration?

[3] The appellant says that if he is sent back to [redacted], he fears he will be
killed by members of a [redacted] organised criminal group (the Syndicate). The
appellant says that Syndicate members blame him for the failure of their criminal
activities in New Zealand.

[4] The appellant contends that the Tribunal’s approach to determining his claims
was incorrect because it engaged in a two-stage assessment consisting of a factual
determination followed by a risk assessment based on limited facts. The appellant’s
position is that the Refugee Convention requires the decision-maker to undertake a
single-stage risk assessment weighing all facts in the mix except those it can
categorically discount. This is because the task of decision-makers assessing refugee
status is not a fact-finding exercise but an assessment of future risk. It is not for the

1
EF v Refugee and Protection Officer [2023] NZCA 372.
Tribunal to make findings about the truth of past events. The appellant says that, by
focusing on a fact-finding exercise, the Tribunal improperly excluded material
information from the risk assessment.

[5] The appellant says that the Tribunal’s approach is out of step with leading
English authority approved by New Zealand courts.2 Karanakaran v Secretary of
State for the Home Department sets out a single-stage test where all evidence
capable of being given any weight at all is considered in an evaluative assessment of
future risk. Evidence can be divided into four categories being:

(1) evidence the decision maker is certain about;

(2) evidence they think is probably true;

(3) evidence to which they are willing to attach some credence, even if they
could not go so far as to say it is probably true; and

(4) evidence to which they are not willing to attach any credence at all.

[6] Only category 4 evidence is discounted. All other evidence forms part of the
risk assessment. The appellant says this approach is correct and was not followed.

[7] The respondent says that the Tribunal did not err. It says that the approach to
risk assessment and refugee status determination is well-settled. When determining
whether a person is at risk of future persecution, the decision-maker must consider
whether evidence is credible and must identify the relevant facts in order to assess
risk. This involves a holistic approach to risk assessment, weighing up all facts and
evidence. The respondent says that the approach to risk assessment applied by the
Tribunal is consistent with both New Zealand and overseas authorities.

Background

[8] EF was granted New Zealand residence in [redacted] because of his marriage
to a New Zealand citizen.

[9] In [redacted], he was sentenced to a lengthy sentence of imprisonment for

his role in the importation of drugs into New Zealand. He was sentenced on the
basis that

2
Karanakaran v Secretary of State for the Home Department [2000] 3 All ER 449 (CA) (UK).
he was effectively a drug mule recruited late in the operation. [redacted] the appellant
delayed and argued with his co-offenders, leading to threats being made against him.

[10] The appellant and his co-offenders were found guilty at trial and sentenced to
various terms of imprisonment. The appellant gave evidence against his co-offenders
at trial, although the evidence was of minimal value given that the entire sequence of
events was captured by police surveillance.

[11] While in prison, the appellant faced further threats from co-offenders who
blamed him for the failure of the operation. He was accused of being an informant. It
was not disputed that the appellant spent the entirety of his sentence in segregation
from the main prison population.

[12] While in custody, the appellant was in contact with his younger brother G in
[redacted]. G informed the appellant, mainly by email, of ongoing threats by the
Syndicate against their older brother H.

[13] The appellant made two claims for refugee and protected person status. Both
claims were based on alleged risk from the Syndicate. The second claim relied on
new evidence namely an allegation that H had been kidnapped and assaulted by the
Syndicate in [redacted].

The first Tribunal decision

[14] The first Tribunal decision is dated [redacted]. The first Tribunal identified
the issues on appeal as twofold. First, whether the appellant’s account was credible
and secondly whether the facts as found met the requisite threshold for risk of
qualifying harm (“a real chance” in terms of the claim for refugee status and “danger
of” in terms of protected person status).

[15] The first Tribunal set out the appellant’s case and then considered the
evidence of [redacted].

[16] [redacted]. He said that H normally lived in [city] but had been on the run for
some years because the Syndicate wished to harm him in revenge for the appellant
causing the loss of their drugs. G said that he first became aware of threats to H
[redacted] but did not advise the appellant. In [redacted], G wrote to the appellant to
tell him that H had been threatened by members of the Syndicate who blamed the
appellant for the loss of the drugs.

[17] The first Tribunal noted that the emails from G advising the appellant of the
alleged threats began shortly after a further court decision. The implication is that the
timing is suspicious and suggests that the threats were manufactured. The timing is
certainly capable of affecting the weight to be attached to G’s evidence.

[18] The first Tribunal said:

[44] In determining whether the appellant is a refugee or a protected

person, it is necessary first to identify the facts against which the assessment
is to be made. That requires consideration of the credibility of the appellant’s
account.

[19] It was accepted that during the offending members of the Syndicate became
angry that the appellant appeared to be retaining the drugs and not handing them
over. There was a flurry of threatening texts and messages received by the appellant.
At that stage, members of the Syndicate had reason to assume that the appellant had
taken the drugs. The first Tribunal said that the threats made at that time must be
seen in that context.

[20] The first Tribunal did not accept that members of the Syndicate continued to
suspect the appellant of taking the drugs. The first Tribunal said that whatever their
initial suspicions, it would have been patently obvious to the Syndicate that the
appellant had nothing to do with the disappearance of the drugs. The first Tribunal
considered that the notes of evidence from the trial revealed that the appellant did not
give any significant evidence against the other Syndicate members.

[21] The first Tribunal accepted that while in custody the appellant was segregated
at his own request from other Syndicate members. The first Tribunal said that the
appellant’s subjective fear of the other men was unsurprising. He was not a member
of the Syndicate. He was recruited by the Syndicate at the last minute and had
received threats from Syndicate members at a time when they assumed him to have
taken the
drugs. His subjective fear and his claims to have heard rumours and gossip while in
prison did not amount to evidence that there was an existing threat.

[22] In respect of the threats against H following the appellant’s convictions, the
first Tribunal did not accept that such threats had been made. In reaching that view it
took account of the fact that the [redacted] threats of harm to the appellant’s family
members were made at a time when the Syndicate believed the appellant had taken
the drugs. Since at least the trial and likely long before that the Syndicate must have
understood the appellant was not to blame. The first Tribunal found there was no
plausible reason why the Syndicate should now be interested in harming H.

[23] The first Tribunal found that the claims by the appellant and G as to the need
for H to go into hiding were implausible. The appellant could not explain why, if it
was safe for H’s wife and children to go to stay [redacted], it was not also safe for H
to do so. Further, the initial claim that H had to abandon his [redacted] business in
[redacted] was undermined by G’s evidence to the refugee status branch as recently
as [redacted] 2018 that [redacted] was continuing to run the business. Nor could G
sensibly explain why the Syndicate would not have forced H’s business to shut by
the simple expedient of scaring off his employees if it was so determined to find and
harm him. A claim that H had narrowly avoided being caught by the Syndicate owed
more to melodrama than to reality and it was significant that H did not give
evidence. The explanation, that G could not contact H because he did not carry a
mobile phone to avoid being traced, was dismissed as nonsensical. The appellant’s
refugee claim had been in existence since [redacted] and there had been ample
opportunity for G to have arranged for H to give evidence.

[24] The first Tribunal said it was implausible that no harm had come to H, even
though the Syndicate had supposedly been pursuing him since 2013. The first
Tribunal also said it was implausible that a syndicate so intent on sending a message
to the appellant would not have sought to harm G who was living, working, and
studying openly in [redacted]. G when giving evidence, asserted vaguely, and the
first Tribunal said unconvincingly, that unknown people had been coming to enquire
about him causing him to leave his home for some weeks. The first Tribunal again
found that claim implausible and said it had no doubt that if G had genuinely
considered that an
international drug syndicate was hunting for him to harm him, he would not have
simply absented himself for a few weeks before returning to the same lifestyle as
before.

[25] The first Tribunal then recorded a heading “the facts as found” and under that
heading listed the following:

[60] The Tribunal finds that the appellant is a [redacted] man from
[redacted]. He holds New Zealand residence [redacted].

[61] The appellant is currently serving [redacted] sentence for his

involvement in the importation [redacted] into New Zealand.

[26] The first Tribunal said that the facts as found established that the appellant is
not a member of any drug syndicate, he was co-opted by one at the last minute to
assist with the collection of drugs. Initial suspicion that he may have betrayed the
Syndicate would inevitably have been dispelled by [redacted] and there was simply
no reason for members of the Syndicate to harbour ill feeling against the appellant
now. The Tribunal said it follows that notwithstanding the common use of violence
by drug gangs [redacted], including in acts of retribution, there were no objective
grounds for believing that the appellant was at risk of serious harm.

[27] The first Tribunal referred to publicity about the trial and said that it is
inherently unlikely that a criminal organisation, which depends on outwitting law
enforcement authorities, would not pay close attention to trials in which members
were prosecuted. The first Tribunal considered the notion that Syndicate members in
[redacted] might still mistakenly blame the appellant to be unreal.

[28] Finally, the first Tribunal said that regard must be had to the passage of time.
The appellant was at that point several years into a lengthy sentence. Given the well-
documented gang violence in [redacted], the first Tribunal said that it could not say
with any confidence that the Syndicate would even exist when the appellant was
eventually released, or that its current members would be alive let alone that they
would remember him or be able to recognise him, a man none of the [redacted]
syndicate members had ever met. The first Tribunal found that there was no
antipathy
towards the appellant in [redacted] but even if there had been, it would be no more
than speculative to find that it would still exist by the time of his release.

[29] The first Tribunal found that while the evidence of the offending and the
appellant’s conviction was obviously correct, his claim to be the subject of ongoing
adverse interest by the Syndicate was not credible. The evidence did not disclose a
real chance of the appellant suffering serious harm arising from breaches of human
rights if he returns to [redacted]. Accordingly, the threshold of risk was not met, and
the appeal was dismissed.

[30] Given those findings, the first Tribunal did not need to address whether there
was a convention reason for the persecution or whether the appellant was excluded
from the protection afforded by the Refugee Convention under art 1F(c).

[31] The conclusions reached by the first Tribunal were that the appellant was:

(a) not a refugee within the meaning of the Refugee Convention;

(b) was not a protected person within the meaning of the

Convention Against Torture; and

(c) was not a protected person within the meaning of the

International Covenant on Civil and Political Rights.

The second Tribunal decision

[32] A second appeal was brought on the basis of new evidence namely an alleged
kidnapping and assault of H following the first Tribunal decision. The second
Tribunal decision was released on [redacted] 2022.

[33] The second Tribunal approached the decision by first setting out the evidence
of H which was the only significant further evidence. H gave evidence that in
[redacted] he began experiencing problems with the Syndicate. At the time, he had a
[redacted] business (not an [redacted] business as the appellant stated in his first
claim). H said he was visited in his shop by a person who informed him that the
appellant was trying to steal drug money. This person was threatening and told H to
call the appellant and get him to return the money. H said that this heralded the
beginning of a three-year campaign of sustained harassment and threats from the
Syndicate which continued during and after the appellant’s trial. H said he was
warned that he and his family would suffer consequences if the Syndicate did not get
their money back. H said that, fearing for his safety and that of his family, he moved
address however, the gang discovered his whereabouts and so he moved again. In
[redacted] took his wife and children to stay [redacted]. Threats by the Syndicate
made business progressively more difficult and by [redacted], H was forced to shut
his business altogether and move from place to place to avoid the Syndicate.

[34] H said that in [redacted] he was kidnapped by [redacted]. He was blindfolded

and driven to a house where he was left alone in a room for a day. He was held for
three days during which he was regularly assaulted and asked where the money was.
On the last day, his foot was deliberately punctured with a metal object so he would
“always remember his motivation to look for the money.” H said that after his release
he went into hiding in a West African country. Since then, he had only returned to
[redacted] once following the death of his father in 2020. He did not attend the
memorial or go to his father’s house located in their family village. said he had
encountered no further problems with the Syndicate since the kidnapping. G had not
encountered problems, something that H put down to the fact that G is considerably
younger than himself and the appellant.

[35] The second Tribunal recorded H’s evidence that the appellant would be in
danger in [redacted] and said that the Syndicate had indicated to him that they were
aware the appellant was still in prison but that proceedings were underway relating
to his eventual return to [redacted]. H gave evidence that he was “running for his
life” at the time of the previous hearing and had cut off most communication with G.

[36] G also gave evidence at the second Tribunal hearing. He said that he became
scared after an unknown person came looking for him at his previous address while
he was away. G left his address for two or three weeks before returning. Three or
four months later, in [redacted], he moved to a new residence. G said there had been
no suspicious visits at his current address, nor had he experienced other problems.
[37] The second Tribunal set out the evidence of the appellant including the fact
that one member of the Syndicate had recently been deported [redacted]. The second
Tribunal accepted that the prison “grapevine” meant that the appellant’s eligibility
for parole would become known to the Syndicate. The appellant gave evidence that
the Syndicate would understand that, if paroled, he would be deported [redacted].
The Syndicate would be able to locate him. The appellant had no means of paying
back the money and said he was at risk of being killed and/or subjected to serious
physical harm.

[38] An undated statement of a lawyer [redacted] attaching a [redacted] affidavit

was relied on by the appellant. The affidavit stated that H visited the lawyer on
[redacted] and that the lawyer sighted injuries to H and particularly to H’s foot. The
lawyer stated that H told him that the injuries were inflicted by members of a drug
cartel who were demanding money from the appellant and that the cartel had
threatened to kill the appellant wherever he is [redacted]. The lawyer confirmed that
H swore an affidavit dated [redacted] regarding that incident and gave his opinion
that it would not be safe for the appellant to return [redacted] because there is a trend
[redacted] for drug cartel members to kill people over drug related matters.

[39] The second Tribunal decision contains a heading “Assessment” under which
it reads:

[31] In determining whether the appellant is a refugee or a protected person

it is necessary to identify the facts against which the assessment is to be
made. That requires consideration of the credibility of the appellant’s
account.

[40] Before making the assessment, the second Tribunal set out its approach to
issues of credibility and its decision-making process. The second Tribunal said that,
contrary to the submissions of both parties at the hearing, the model taken in
New Zealand refugee status determinations is a single-stage assessment consistent
with the approach in Karanakaran.

[41] The second Tribunal said that what the appellant’s counsel characterised “as
an erroneous or impermissible first stage”, being the “credibility” and “findings of
facts” sections, is simply the mechanism by which the Tribunal identifies, with
reasons, those portions of the evidence about which it has no real doubt. In terms of
the Karanakaran evidential schematic, this part of the process simply functions to
highlight those portions of the evidence to which it is “not willing to attach any
credence at all” and thus are to be given no weight in the forward-looking
assessment of risk.3 These are the category 4 factors on a Karanakaran analysis.4

[42] The Tribunal said that the structure of its decisions did not convert the
underlying process into a “two-stage process.” Rather it reflected that this element of
the decision-making process engages a duty to give adequate reasons just as much as
those elements relating to the assessment of future risk. All assertions of fact not
identified as being “not credible” are accepted although the assertions will vary in
their relevance to the asserted risk depending on the claim advanced. Some will be
accepted because they are demonstrably true. Others will be accepted because the
benefit of the doubt is applied (for example where the appellant’s oral evidence is
plausible but lacks corroboration). In recording the facts, the second Tribunal said
that it typically summarises the key facts relevant to the risk enquiry.

[43] The second Tribunal said that it does not and never has approached credibility
assessments on the basis of there being an applicable “standard of proof.” It has
consistently recognised that its function and task is fundamentally evaluative. It is
for this reason that its credibility assessments do not distinguish in the “facts as
found” section between the remaining Karanakaran categories. In the “facts as
found” section of any tribunal decision, there will be differing degrees of certainty
the decision-maker has in relation to those facts but, for good reason, no distinction
is made between them.

[44] The second Tribunal said that any differentiation between the accepted facts
relates to their relative significance in the forward-looking risk assessment, not the
underlying status as accepted facts upon which that assessment rests. In other words,
they are weighted in terms of risk, not relative certainty. Each is an accepted fact and
is regarded as equally true.

3
Kaja v Secretary of State for the Home Department [1995] ImmAR 1.
4
Karanakaran v Secretary of State for the Home Department, above n 2, at 459
[45] The second Tribunal said that in many ways the “benefit of doubt” principal
is a misnomer. In truth, it is an expression of uncertainty. It only comes into play
when there are factors which cause the decision-maker to have concern about
categorising the asserted fact as one for which they have “no real doubt.” In
recognising the dangers of making “the wrong mistake” uncertainty is resolved in the
appellant’s favour.

[46] The second Tribunal quoted DS (Iran) where it was said:5

[1] The process of refugee status determination has aptly been described
as inherently one of the most challenging in the legal world; … . The process
involves evaluating future risks, not judging the more believable of two
competing accounts of past events. It is not generally possible to verify a
claim by conducting investigations in the country of origin to test the
assertions made by the claimant; …. Typically, the only witness of fact will
be the claimant, sometimes traumatised, and with cultural traits and life
experiences often very different from those of the decision-maker. Speaking
extra- judicially, Sir Stephen Sedley, former Lord Justice of Appeal in the
United Kingdom, correctly observed that the process involves making ‘a
possible life- and-death decision extracted from shreds of evidence and
subjective impressions”, in a jurisdiction which “has neither the falsifiability
of a science nor the completeness of an art’.

[47] The second Tribunal said that the realities are well understood by it and that
its approach to both credibility and the assessment of risk reflects this.

[48] In assessing the credibility of evidence in the second claim, the second
Tribunal said that the striking feature of the claim was the emergence in [redacted] of
an incident in which H claimed that he was kidnapped and injured in the foot with a
metal implement in what was described as a “final warning.” The second Tribunal
regarded it as significant that, in the decision declining the first appeal one of the
reasons the first Tribunal gave for rejecting the credibility of the claim was precisely
because H had not been harmed by the Syndicate. The first Tribunal’s decision in
respect of the appellant’s first claim was issued [redacted], a mere three months
before the claimed kidnapping incident occurred. This raised a concern that the
incident was manufactured in response to the first Tribunal’s finding.

5
DS (Iran) [2016] NZIPT 800788.
[49] While there was no doubt that H had suffered a puncture wound to the sole
of his foot, the location and nature of the injury suggested that it was far more likely
to have been caused by H stepping on something sharp by accident. The credibility
issues, when aggregated, left the second Tribunal with no uncertainty that the claim
of alleged kidnapping and mistreatment of H was untrue. In respect of G, the second
Tribunal found his evidence was measured and credible but did little to advance the
appellant’s assertion that there was a real chance that he would be subjected to
retaliatory action by the Syndicate if returned [redacted].

[50] The second Tribunal found no reason to depart from the first Tribunal’s
findings that, although the Syndicate may have thought at the time of the seizure of
the drugs [redacted], that the appellant was to blame, the trial would have made it
very clear this was not the case.

[51] The second Tribunal then set out the facts in a section of the decision headed
“The facts as found” and said:

[66] The Tribunal adopts the findings of fact made by the first panel,
namely that the appellant is [redacted] man from [country]. He holds New
Zealand residence [redacted].

[67] The appellant is currently service a [redacted] sentence for his

involvement in the importation of [redacted] into New Zealand. [redacted].

[68] The appellant has two brothers. [H] resides in a [redacted] where he
is assisting a friend for his employment. This brother’s family live with
[redacted]. [O] has recently [redacted] and lives in [city]. [redacted], an
unknown person visited his former residence and, as a precautionary
measure, he changed his place of residence.

[52] The second Tribunal accepted that [redacted] was experiencing a prevalence
of drug syndicates and subsequent anti-social impacts.

[53] Having discounted the new evidence of H, the second Tribunal found that the
appellant was not entitled to be recognised as a refugee or as a protected person.

The issues

[54] The question of law to be decided comprises two distinct questions:

 (a) Did the Tribunal err in its approach to risk assessment?

(b) If it did, did the Tribunal, as a result, improperly exclude material

information?

Discussion

[55] The appellant submits that the correct approach is that taken in Karanakaran
and says that the way in which both Tribunals approached the decision-making
process was not in accordance with Karanakaran.

[56] Karanakaran confirmed the approach taken by the

Immigration Appeal Tribunal in an earlier decision Kaja v Secretary of State for the
Home Department.6 The Court in Karanakaran summarised the effect of Kaja
saying:7

The majority of the tribunal considered that the question they had to decide
was whether the assessment of an asylum case was a two-stage process or a
one-stage process. They considered that it was a one-stage process. The task
of the decision-maker was to assess, to a reasonable degree of likelihood,
whether the applicant’s fear of persecution for a convention reason was well-
founded. It might be that there were parts of the evidence which on any
standard were to be believed or not to be believed. Of other parts, the best
that might be said of them was that they were more likely than not. Of other
parts it might be said that there was a doubt. The need to reach a decision on
whether an appellant had made his case to a reasonable degree of likelihood,
arose only on the ultimate evaluation of the case, when all the evidence and
the varying degrees of belief or disbelief were being assessed.

The majority considered that if there was a first stage (proof of present and
past facts) followed by a second stage (assessment of risk) then any
uncertainties in the evidence would be excluded at the second stage, and that
this could not be right. In those circumstances, they considered that the
introduction of an intervening stage was simply an unnecessary complexity.

What they decided was that when assessing future risk decision-makers may
have to take into account a whole bundle of disparate pieces of evidence:

(1) evidence they are certain about;

(2) evidence they think is probably true;

6
Kaja v Secretary of State for the Home Department, above n 3.
7
Karanakaran v Secretary of State for the Home Department, above n 2, at 458–459.
 (3) evidence to which they are willing to attach some credence, even if they
could not go so far as to say it is probably true;

(4) evidence to which they are not willing to attach any credence at all.

[57] The effect of such an assessment is that the decision-maker would not
exclude any evidence except category 4 evidence. Category 3 evidence, although
unproven on the balance of probabilities, would remain relevant to the risk
assessment.

[58] The Court in Karanakaran summarised the similar approach taken by a line
of Australian authorities before Brooke LJ concluded with the agreement of the rest
of the Court:8

This approach does not entail the decision-maker (whether the Secretary of
State or an adjudicator or the Immigration Appeal Tribunal itself) purporting
to find ‘proved’ facts, whether past or present, about which it is not satisfied
on the balance of probabilities. What it does mean, on the other hand, is that
it must not exclude any matters from its consideration when it is assessing
the future unless it feels that it can safely discard them because it has no real
doubt that they did not in fact occur (or, indeed, that they are not occurring at
present). Similarly, if an applicant contends that relevant matters did not
happen, the decision-maker should not exclude the possibility that they did
not happen (although believing that they probably did) unless it has no real
doubt that they did in fact happen.

For the reasons much more fully explained in the Australian cases, when
considering whether there is a serious possibility of persecution for a
convention reason if an asylum-seeker is returned, it would be quite wrong
to exclude matters totally from consideration in the balancing process simply
because the decision-maker believes, on what may sometimes be somewhat
fragile evidence, that they probably did not occur.

[59] Sedley LJ observed:9

No probabilistic cut-off operates here: everything capable of having a

bearing has to be given the weight, great or little, due to it.

[60] The Australian authorities which informed the Court in Karanakaran include
Minister for Immigration and Multicultural Affairs v Rajalingam,10 and Abebe v
Commonwealth of Australia.11

8
At 469–470.
9
At 479.
10
Minister for Immigration and Multicultural Affairs v Rajalingam [1999] FCA 719, (1999) 93
FCR 220.
11
Abebe v Commonwealth of Australia (1999) 197 CLR 510.
[61] More recent Australian authorities have again confirmed Karanakaran and
Abebe. In BAO16 v Minister for Immigration and Border Protection the Court cited
with approval the remarks of Brooke LJ in Karanakaran quoted above and found
that a decision-maker had improperly excluded information from the risk
assessment.12 The Court said:13

The decision-maker must still consider the extent to which it is likely that the
claimed event occurred in assessing whether the claimant has a well-founded
fear of persecution, and the decision-maker cannot avoid “reasonable
speculation” about the chances of persecution when the claimant’s material
is considered as a whole. In so doing, and as part of the core task of
evaluating and weighing the evidence, the decision-maker must consider
how much weight should be given to these matters, having regard to the
other claims and the evidence in the case. The assessor in this case
apparently failed to appreciate that, in considering the issue of well-founded
fear, hers was essentially an evaluative task.

[62] The Australian jurisprudence on the Refugee Convention uses the word
“speculation” to describe a legitimate exercise falling short of fact finding.

[63] In the leading judgment of the High Court of Australia Minister for
Immigration and Ethic Affairs v Wu Shan Liang Kirby J said:14

… the decision-maker must not, by a process of factual findings on

particular elements of the material which is provided, foreclose reasonable
speculation upon the chances of persecution emerging from a consideration
of the whole of the material. Evaluation of chance, as required by Chan
cannot be reduced to scientific precision. That is why it is necessary,
notwithstanding particular findings, for the decision-maker in the end to
return to the question: “What if I am wrong”? 15 Otherwise, by eliminating
facts on the way to the final conclusion, based upon what seems “likely” or
“entitled to greater weight”, the decision-maker may be left with nothing
upon which to conduct the speculation necessary to the evaluation of the
facts taken as a whole, in so far as they are said to give rise to a “real
chance” of persecution.’

[64] In Wu Shan Liang Kirby J concluded:16

Ultimately, the question is whether the [decision-maker] allowed her mind to

consider all the relevant possibilities by looking back at the entirety of the
material placed before her and considering it against a test of what the “real”,

12
BAO16 v Minister for Immigration and Border Protection [2018] FCA 1463.
13
At [91].
14
Minister for Immigration and Ethic Affairs v Wu Shan Liang (1196) 185 CLR 259 at 507–508.
15
Guo Wei Rong v Minister for Immigration and Ethnic Affairs (1996) 135 ALR 421 at 441 per
Einfeld J.
16
Minister for Immigration and Ethic Affairs v Liang, above n 14, at 508.
 as distinct from fanciful, “chances” would bring if the applicant were
returned to China.

[65] In Minister for Immigration in Multicultural Affairs v Rajalingam the Court

adopted the following finding by Drummond J in Thanh Phat Ma v Billings:17

… unless the decision-maker can dismiss as unfounded factual assertions

made by the applicant, the decision-maker should be alert to the importance
of considering whether the accumulation of circumstances, each of which
possesses some probative cogency, is enough to show, as a matter of
speculation, a real chance of persecution, even though no one circumstance,
considered by itself, is sufficient to raise that prospect.

[66] Brooke LJ in Karanakaran, summarising the principles derived by Sackville

J in Rajalingam said:18

There may be circumstances in which a decision-maker must take into

account the possibility that alleged past events occurred even though it finds
that these events probably did not occur. The reason for this is that the
ultimate question is whether the applicant has a real substantial basis for his
fear of future persecution. The decision-maker must not foreclose reasonable
speculation about the chances of the future hypothetical event occurring.

[67] The New Zealand approach and whether it amounted to an impermissible

two-stage analysis was considered but not decided by the Court of Appeal in
DY (Pakistan) v Refugee and Protection Officer.19

[68] The principles set out in Karanakaran have been cited with approval by the
Tribunal and by New Zealand Courts.20 References to Karanakaran have however
tended to focus on other aspects of the decision particularly the rejection of a balance
of probabilities test for the determination of relevant risk factors. Karanakaran is
authority for the proposition that the benefit of the doubt in respect of refugee status
determinations operates to resolve doubt in favour of a claimant. This is
uncontroversial in New Zealand and appears to be accepted as correct and applied by
the Tribunal. Where the Tribunal has a doubt about whether evidence can be placed

17
Thanh Phat Ma v Billings (1996) 71 FCR 431 at 436.
18
Karanakaran v Secretary of State for the Home Department, above n 2, at 468.
19
DY (Pakistan) v Refugee and Protection Officer [2021] NZCA 522.
20
Jiao v Refugee Status Appeals Authority [2003] NZAR 647 (CA) endorsing Refugee Appeal No
72668/01 [2002] NZAR 649 which cited Karanakaran v Secretary of State for the Home
Department, above n 2.
in category 4 of the Karanakaran categories (and therefore discounted) the benefit of
that doubt should be resolved in favour of the applicant.

[69] The second Tribunal purported to adopt the Karanakaran approach and
agreed it is the correct approach. The second Tribunal said that it did apply a single-
stage test consistent with Karanakaran and that the structure of the decision was
simply a convenient way to organise its reasons.

[70] A reading of both decisions does not sit easily with the contention that the
Karanakaran approach was followed. Both Tribunals appear to completely dismiss
the risk that Syndicate members remain angry with the appellant. The Tribunals
appeared to find that members of the Syndicate would inevitably take a logical and
reasonable approach to blame relying on the evidence at trial and would realise that
the appellant was not to blame for the interception of the drugs. It seems optimistic,
however, to entirely dismiss the risk that Syndicate members would continue to hold
an adverse view of the appellant. The Tribunals had no way of knowing the size of
the Syndicate or how connected all members were to each other or indeed how
rational and logical individual members of the Syndicate might be. The Tribunals did
not factor in the risk of misinformation.

[71] This is not to say that the risk of the Syndicate members continuing to hold
an adverse view of the appellant is in any way decisive. That risk must be weighed
against the passage of time, the fact that [redacted] based syndicate members never
met the appellant, and the measures the appellant himself can take to ensure he
remains distant from drug syndicates. The problem is that the Tribunal appears to
have discounted entirely a risk that should have been weighed in the balance.

[72] Part of the problem is the way the Tribunal structures its decisions. It may
well be correct that the Tribunal endeavours to take an evaluative and holistic
approach to risk assessment but that is not apparent from the decisions. Rather the
decisions are structured in such a way that it appears that the Tribunal has made
factual findings discounting risk factors and has then assessed future risk with no
reference to risk factors which should not have been entirely discounted. Where the
Tribunal has found that a particular risk factor carries no weight at all, it is proper to
disregard it, but not
all the risk factors which are absent from the final analysis appear to fall into that
category.

[73] It is important that refugee status decisions are structured in such a way that
the decision-maker cannot dismiss risk factors that are difficult to resolve. It is also
important that the decisions are set out in such a way that the actual reasoning
process is apparent. This provides for a robust reasoning process and ensures that all
relevant risk factors are properly weighed as part of the ultimate enquiry.

[74] I have little difficulty in concluding that the Karanakaran approach is the
correct approach. This aligns the New Zealand approach to that of both Australia and
the United Kingdom. In Tamil X v Refugee Status Appeals Authority, the Court of
Appeal said that whether a person qualifies as a refugee should not vary depending
upon the jurisdiction in which the person makes the claim.21 As an international
treaty the Refugee Convention should receive a uniform interpretation across
jurisdictions which are party to it.22

[75] There is also no question that the Tribunal is entitled to make credibility
findings. Under s 135 of the Immigration Act it is the responsibility of a claimant to
establish their claim for recognition as a refugee or a protected person. Section
137(5) of the Immigration Act specifically states that a Refugee and Protection
officer in determining the status of a claimant “may make findings of credibility or
fact.”

[76] The correct approach to credibility assessments in the context of applications

for refugee or protected person status is well-settled in New Zealand. 23 Because there
are real limitations upon an applicant in establishing a refugee claim through
supporting evidence, a generous approach is called for, which includes the benefit of
a doubt principle, by which, if an applicant’s account for refugee status is credible,
the applicant should be given the benefit of the doubt even in the absence of
corroborative material.24 Statements which are sought to be relied on must not run
counter to

21
Tamil X v Refugee Status Appeals Authority [2009] NZCA 488, [2010] 2 NZLR 73 at [151].
22
At [151].
23
AR v Immigration and Protection Tribunal [2017] NZHC 2039, [2017] NZAR 1524 at [10].
24
BV v The Immigration and Protection Tribunal [2014] NZCA 594, [2015] NZAR 139 at [6]; and
Jiao v Refugee Status Appeals Authority, above n 20.
generally known facts. Even if the broad framework of an applicant’s contentions is
coherent, detailed inconsistencies destructive of credibility should not be
disregarded.25 There is nothing objectionable in rejecting an applicant’s claim by
reference to various inconsistencies.

[77] In Attorney-General (Minister of Immigration) v Tamil X the Supreme Court

held that the inapplicability of rules of evidence gives the authority a broad
discretion as to what material it obtains and uses in consideration of a claim. 26 A
realistic and careful approach to that material must be taken having regard to the
evidential gaps and other difficulties that refugee claimants face in making their
claims. A legitimate and important consideration will often be whether the
applicant’s evidence is credible and plausible.27

[78] The question in the present case is not whether the Tribunal can make
credibility findings nor is it whether the credibility finding in respect of H’s evidence
about the alleged kidnapping was open to the Tribunal. Both of those questions can
be easily answered in the affirmative. The question concerns the correct
methodology to be applied to the decision-making process.

[79] In Karanakaran Brooke LJ dealt with methodology to be applied and said:28

I express no view on the merits of that decision. That is, and must be, a
matter for the judgment of the members of that experienced specialist
tribunal, and unless they have committed some error of law this court will
not interfere with their judgment. What is relevant in the present context is
the methodology they adopted. Unless something is so trivial that even on a
cumulative assessment it would be bound to carry no weight, or the decision-
maker has no real doubt that it is entitled to discard some point from its
consideration altogether, it would be wrong to eliminate that point
completely.

[80] The correct approach is to identify the evidence relied on by both parties and
to determine whether any evidence falls into category 4 of Karanakaran. That may
well involve credibility and plausibility assessments. Evidence which falls into
category 4 carries no weight and can be discounted. The tribunal must assess risk

25
Cao v Immigration and Protection Tribunal [2014] NZHC 259 at [18].
26
Attorney-General (Minister of Immigration) v Tamil X [2010] NZSC 107, [2011] 1 NZLR 721 at
[44].
27
At [44].
28
Karanakaran v Secretary of State for the Home Department, above n 2, at 471.
taking into account all of the remaining evidence. Matters which carry little weight
individually may cumulatively establish that the appellant’s fear of persecution is
well founded. Equally matters which mitigate risk may cumulatively operate to
establish that there is no real risk or that the risk can be adequately managed in the
appellant’s home country.

[81] An example in the present case would be the evidence that the appellant was
previously threatened by members of the Syndicate. That carries weight but the
weight must be balanced against the fact that the appellant did not in fact betray the
Syndicate and that should now be known to the Syndicate. That must in turn be
balanced against the risk of misinformation and incorrect information being relied on
by Syndicate members. The previous threats must be balanced against the passage of
time, the fact that the appellant is not personally known to the members of the
Syndicate and the lack of any harm or approach to G or other members of the family
who would be relatively easy to locate.

[82] This evaluation of weight was discussed by Brooke LJ in Karanakaran who

said:29

… when assessing the future, the decision-maker is entitled to place greater

weight on one piece of information rather than another.

This balancing exercise may necessarily involve giving greater weight to

some considerations than to others, depending variously on the degree of
confidence the decision-maker may have about them, or the seriousness of
their effect on the asylum-seeker’s welfare if they should, in the event, occur.

[83] Brooke LJ approved the Tribunal’s technique in Sayandan’s case of

evaluating both the likelihood of a risk eventuating and the seriousness of the
consequences if it were to eventuate. 30 Brooke LJ said it was also correct for the
Tribunal to assess the cumulative effect of the matters it was considering, particularly
if there was a likelihood that they would all affect the applicant at the same time. 31
Some evidence will carry much weight, some will not. Some will carry no weight at
all. The decision-

29
At 470.
30
At 471 referring to Sayandan v Secretary of State for the Home Department IAT, 5 March 1998
(UK).
31
At 471–472.
maker must evaluate the cumulative weight of the evidence to determine whether the
applicant meets the criteria for recognition as a refugee or protected person.32

[84] The Tribunal should be able to articulate its reasons for placing little weight
on some parts of the evidence and more weight on others. It is important to
remember that it is not necessary to reject the entirety of a witness’s evidence. A
witness may exaggerate a threat, but some weight may be placed on the evidence
that there is a threat, that weight being reduced by apparent exaggeration.

[85] It is entirely unclear from the decisions what weight if any was attached to
the risk of deported co-defendants continuing to bear a grudge against the appellant
despite the trial evidence. The Tribunal needed to ask itself what if the co-defendants
still thought the appellant responsible in some way for what happened? What if they
hold an irrational view about that? What is the risk they would seek him out? What
is the likelihood they would have a chance encounter and if they did would they
recognise him? This is the sort of speculation about the chances of hypothetical
future risk envisaged by the Australian authorities. All of those questions may be
answered by competing factors that mitigate risk, but such questions need to be
considered.

[86] The format of both decisions supports the appellant’s contention that the
Tribunal engaged in a two-stage approach whereby it considered and discounted
evidence of past facts before moving onto an assessment of risk based on a narrow
view of the facts.

[87] The second Tribunal said that the “credibility” and “findings of facts”
sections, are simply the mechanism by which the Tribunal identifies, with reasons,
the portions of the evidence about which it has no real doubt. That is difficult to
reconcile with the items listed under the “facts as found” heading in both decisions.
In the first decision, the only facts listed were that the appellant is a [redacted] from
[country] who holds New Zealand residence [redacted]. [redacted], and the appellant
is currently serving a lengthy sentence for his involvement in the importation of
drugs into New Zealand. In the second decision the second Tribunal adopted the
facts as found by the first Tribunal and added information about the appellant’s
brother G receiving a visit from

32
See James A Sweeney “Credibility, Proof and Refugee Law” (2009) 21 IJRL 700.
an unknown person which caused him to change his place of residence as a
precautionary measure.

[88] Other facts about which the Tribunal can have had no doubt are not listed in
the “facts as found section.” This includes the fact that the appellant was subjected to
threats from the Syndicate, he was segregated for his entire prison sentence, and his
co-defendants have been deported to [redacted]. These facts are mentioned elsewhere
but if the “facts as found” section is a way to set out the facts about which the
Tribunal has no doubt, it does not do that. The structure of the decisions is such that
a two-stage process is encouraged even if that is not intended.

[89] It is impossible to avoid the conclusion that the Tribunal excluded material
information from consideration. The most obvious example is consideration of the
risk that Syndicate members would continue to harbour ill-feeling toward the
appellant notwithstanding the evidence at trial. The Tribunal appeared to completely
disregard that possibility. The Tribunal assumed a logical and rational approach by
all members of a drug syndicate. Had the analysis been carried out in the way
suggested in Karanakaran, the possibility of ongoing ill-will by the Syndicate
towards the appellant could not be completely discounted. It may well be that the
evaluation of the weight to be attached to that risk would result in the same outcome,
but it should not have been excluded from consideration.

[90] The second Tribunal differed from the first in its view of G’s credibility.
There was no discussion of the evidence that G became aware of ongoing threats
between [redacted]. Again, the weight of that evidence might have been considerably
affected if the information came to G from H. But that was not considered. If the
information did come from H that would suggest that H was complaining of ongoing
threats from [redacted]. That might properly have led the Tribunal to consider
whether any credence should attach to H’s evidence even if the later kidnapping
claim was entirely discredited.

[91] In the present case, the Tribunal did err in its approach to risk assessment by
failing to clearly distinguish evidence to which no credence at all should be attached
from evidence which did have some credence. The Tribunal also erred by failing to
clearly evaluate the weight to be applied to evidence to which it attached little
credence but nevertheless remained in the mix.

[92] It follows from this discussion that the question posed by the Court of Appeal
must be answered in the affirmative. The Tribunal did err in its approach to risk
assessment and, as a result, it did improperly exclude material information from
consideration.

[93] There was a suggestion that this Court should determine the substantive issue.
I agree with Wylie J who said in the unsuccessful High Court leave application that it
is far from clear that a different approach would lead to a different outcome for EF.33
That is a matter best determined by the Tribunal whose members have specialist
expertise and experience in this area.

Result

[94] I answer the question of law in the affirmative and I remit the matter to the
Tribunal for rehearing.

Wilkinson-Smith J

33
EF v Refugee and Protection Officer [2022] NZHC 3371 at [40(e)].