![]() |
[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback] [DONATE] | |||||||||
Upper Tribunal (Immigration and Asylum Chamber) |
||||||||||
PLEASE SUPPORT BAILII & FREE ACCESS TO LAW
To maintain its current level of service, BAILII urgently needs the support of its users.
Since you use the site, please consider making a donation to celebrate BAILII's 25 years of providing free access to law. No contribution is too small. If every visitor this month gives just £5, it will have a significant impact on BAILII's ability to continue providing this vital service.
Thank you for your support! | ||||||||||
You are here: BAILII >> Databases >> Upper Tribunal (Immigration and Asylum Chamber) >> MST and Others (national service – risk categories) Eritrea CG [2016] UKUT 443 (IAC) (7 October 2016 ) URL: https://www.bailii.org/uk/cases/UKUT/IAC/2016/443.html Cite as: [2016] UKUT 443 (IAC) |
[New search]
[Context]
[View without highlighting]
[Printable PDF version]
[Help]
Upper Tribunal
(Immigration and Asylum Chamber)
MST and Others (national service – risk categories) Eritrea
CG [2016]
UKUT 443 (IAC)
Last updated on: 23 December 2016
THE IMMIGRATION ACTS
Heard at Field House |
Decision & Reasons Promulgated |
On 2, 3, 6, 7, 8, 10 & 20 June 2016 |
|
|
|
Before
UPPER TRIBUNAL JUDGE H H STOREY
UPPER TRIBUNAL JUDGE McWILLIAM
Between
MST
MYK
AA
(ANONYMITY DIRECTION made)
Appellants
And
THE SECRETARY OF STATE FOR THE HOME DEPARTMENT
Respondent
And
UNITED NATIONS HIGH COMMISSIONER FOR REFUGEES
Intervener
Representation:
For the Appellants: Mr S Knafler, QC, Ms A Benfield, Mr T Hussain, instructed by Immigration Advice Service, Fountains Solicitors and Roelens Solicitors
For the Respondent: Mr B Rawat, Mr R Harland, instructed by Government Legal Department
For the Intervener: Ms L Dubinsky and Mr T Pascoe instructed by Baker & McKenzie LLP/UNHCR
Country guidance
1. Although reconfirming
parts of the country guidance given in MA (Draft evaders – illegal
departures – risk) Eritrea
CG [2007] UKAIT 59 and MO (illegal exit –
risk on return)
Eritrea
CG [2011] UKUT 190 (IAC), this case replaces that
with the following:
2. The Eritrean system of military/national service remains indefinite and since 2012 has expanded to include a people’s militia programme, which although not part of national service, constitutes military service.
3. The age limits for national service are likely to remain the same as stated in MO, namely 54 for men and 47 for women except that for children the limit is now likely to be 5 save for adolescents in the context of family reunification. For peoples’ militia the age limits are likely to be 60 for women and 70 for men.
4. The categories of lawful exit have not significantly changed since MO and are likely to be as follows:
(i) Men aged over 54
(ii) Women aged over 47
(iii) Children aged under five (with some scope for adolescents in family reunification cases
(iv) People exempt from national service on medical grounds
(v) People travelling abroad for medical treatment
(vi) People travelling abroad for studies or for a conference
(vii) Business and sportsmen
(viii) Former freedom fighters (Tegadelti) and their family members
(ix) Authority representatives in leading positions and their family members
5. It continues to be the
case (as in MO) that most Eritreans who have left Eritrea
since 1991 have done so illegally. However, since there are viable, albeit still limited,
categories of lawful exit especially for those of draft age for national
service, the position remains as it was in MO, namely that a person
whose asylum claim has not been found credible cannot be assumed to have left
illegally. The position also remains nonetheless (as in MO) that if such
a person is found to have left
Eritrea
on or after August/September 2008, it
may be that inferences can be drawn from their health history or level of
education or their skills profile as to whether legal exit on their part was
feasible, provided that such inferences can be drawn in the light of adverse
credibility findings. For these purposes a lengthy period performing national
service is likely to enhance a person’s skill profile.
6. It remains the case (as in MO) that failed asylum seekers as such are not at risk of persecution or serious harm on return.
7. Notwithstanding that the round-ups (giffas) of suspected evaders/deserters, the “shoot to kill” policy and the targeting of relatives of evaders and deserters are now significantly less likely occurrences, it remains the case, subject to three limited exceptions set out in (iii) below, that if a person of or approaching draft age will be perceived on return as a draft evader or deserter, he or she will face a real risk of persecution, serious harm or ill-treatment contrary to Article 3 or 4 of the ECHR.
(i) A person who is likely to be perceived as a deserter/evader will not be able to avoid exposure to such real risk merely by showing they have paid (or are willing to pay) the diaspora tax and/have signed (or are willing to sign) the letter of regret.
(ii) Even if such a person may avoid punishment in the form of detention and ill-treatment it is likely that he or she will be assigned to perform (further) national service, which, is likely to amount to treatment contrary to Articles 3 and 4 of the ECHR unless he or she falls within one or more of the three limited exceptions set out immediately below in (iii).
(iii)
It remains the case (as in MO) that there are persons likely not to face
a real risk of persecution or serious harm notwithstanding that they will be
perceived on return as draft evaders and deserters, namely: (1) persons whom
the regime’s military and political leadership perceives as having given them
valuable service (either in Eritrea
or abroad); (2) persons who are trusted
family members of, or are themselves part of, the regime’s military or
political leadership. A further possible exception, requiring a more case
specific analysis is (3) persons (and their children born afterwards) who fled
(what later became the territory of)
Eritrea
during the War of Independence.
8. Notwithstanding that
many Eritreans are effectively reservists having been discharged/released from
national service and unlikely to face recall, it remains unlikely that they
will have received or be able to receive official confirmation of completion of
national service. Thus it remains the case, as in MO that “(iv) The
general position adopted in MA, that a person of or approaching draft
and not medically unfit who is accepted as having left Eritrea
illegally is
reasonably likely to be regarded with serious hostility on return, is
reconfirmed, subject to limited exceptions…”
9. A person liable to
perform service in the people’s militia and who is assessed to have left
Eritrea
illegally, is not likely on return to face a real risk of persecution
or serious harm.
10. Accordingly, a person whose asylum claim has not been found credible, but who is able to satisfy a decision-maker (i) that he or she left illegally, and (ii) that he or she is of or approaching draft age, is likely to be perceived on return as a draft evader or deserter from national service and as a result face a real risk of persecution or serious harm.
11. While likely to be a rare case, it is possible that a person who has exited lawfully may on forcible return face having to resume or commence national service. In such a case there is a real risk of persecution or serious harm by virtue of such service constituting forced labour contrary to Article 4(2) and Article 3 of the ECHR.
12. Where it is specified above that there is a real risk of persecution in the context of performance of military/national service, it is highly likely that it will be persecution for a Convention reason based on imputed political opinion.
TABLE OF CONTENTS
GLOSSARY
INTRODUCTION Paragraphs 1
The true meaning of country guidance 4
The country guidance issues 10
Existing country guidance 15
Strasbourg cases on Eritrea
23
Decisions
of national courts and tribunals on Eritrea
24
The legal issues 25
The appellants 26
Procedural matters 28
A. EVIDENCE AND SOURCES 30
1. Background evidence 31
UNHCR Eligibility Guidelines for Assessing the 31
International Protection Needs for
Asylum-seekers from Eritrea
, 20 April 2011
Eritrean Ministry of Information publication: “UNHCR 32
Eligibility Guidelines: Factual Findings or
Recycled Reformation”, 17 December 2015
Danish Fact-Finding Mission (DFFM) Reports, 35
25 November and 16 December 2014
Landinfo, 23 March 2015 and 16 April 2015 and May 2016 45
UN Commission of Inquiry Reports 5 June 2015 and 8 June 2016 49
EASO Country of Origin Information Report, 57
Eritrea
Country Focus, May 2015
US State Department Country Reports on Human Rights Practices, 59
chapter on Eritrea
, 2014 and 2015
Amnesty International, Report on AA, 22 September 2015, 63
and the “Just Deserters” Report, December 2015
Lifos
Reports on Eritrea
, 23 November and 15 December 2015 74
Swiss Visits, January 2016 and March 2016 77
Human Rights Watch Reports 80
UK government materials 83
The new versions of Home Office CIGs, 4 August 2016 102
Academics and journalists 125
Bisha Mines materials 142
2. Expert evidence of Professor Kibreab (PK) in summary form 147
B. ASSESSMENT: THE GENERAL ISSUES 148
1. Law
The relevance of existing country guidance 148
The status of UNHCR Eligibility Guidelines and position papers 155
The status of experts in country guidance cases 159
2. Methodology and Sources
General Observations 161
Anonymity of sources 163
Fact-finding mission reports: general 165
Fact-Finding Missions and the Eritrean context 168
The Danish Fact Finding Mission (DFFM) Report 172
The UK Fact Finding Mission (UKFFM) materials 192
The two Amnesty International Reports 202
(AI Report on AA and “Just Deserters”), the two
UNCOI Reports of 2015 and 2016 and the
witness statement from Elizabeth Chyrum
The UNCOI Reports 2015 and 2016 209
Witness statement of Elizabeth Chyrum 224
The Home Office Country Information and Guidance (CIG) 225
publications on 4 August 2016
C. ASSESSMENT OF THE EVIDENCE OF PROFESSOR KIBREAB 228
D. FINDINGS ON MAIN GENERAL ISSUES 241
The general situation 242
National Service 247
People’s Militia 260
Submissions 265
Our assessment 280
Enforcement and Punishment 280
Conditions 284
Eligibility and duration 287
Exemptions 291
Demobilisations/ discharges and release 297
Eligibility for national service and exit visas 308
Submissions 311
Our assessment 316
The 2 per cent tax and letter of regret 329
Submissions 329
Our assessment 333
Failed Asylum Seekers 335
Illegal exit by those perceived on return to be draft-evaders or deserters 338
Submissions 338
Our assessment 344
Forcible Returns 357
Submissions 358
Our assessment 360
Draft evaders and deserters 368
National service as slavery or servitude or forced labour 371
Article 4 and Article 3 376
Article 4: the legal framework 378
The threshold test 390
The Eritrean Context: the ILO background 399
Slavery and servitude: our assessment 402
Forced or compulsory labour: our assessment 416
Conclusions 431
E. ASSESSMENT: THE APPELLANTS 433
APPENDICES
APPENDIX I: THE EVIDENCE RELATING TO THE APPELLANTS I
APPENDIX II: ERROR OF LAW DECISIONS FOR APPELLANTS II
APPENDIX III: THE EVIDENCE OF PROFESSOR KIBREAB III
APPENDIX IV: SCHEDULE OF BACKGROUND EVIDENCE IV
GLOSSARY
AI Amnesty International
AIR Asylum Interview Record
CEACR Committee of Experts on the Application of Conventions and Recommendations
CIG Country Information and Guidance
CIPU Country Information and Policy Unit
CMR Case Management Review
COI Country of Origin Information
COIS Country of Origin Information Service
CPIT Country Policy and Information Team
CSLT Country Specific Litigating Team
DFFM Danish Fact-Finding Mission Reports
DIS Danish Immigration Service
DL Discretionary Leave
EASO European Asylum Support Office
ECtHR European Court of Human Rights
FCO Foreign and Commonwealth Office
FLC Forced Labour Convention
HRW Human Rights Watch
HRCE Human
Rights Concern – Eritrea
IAGCI Independent Advisory Group on Country Information
ICC International Criminal Court
ICCPR International Covenant on Civil and Political Rights
ICG International Crisis Group
ICRC International Committee of the Red Cross
ICTY The International Criminal Tribunal for the former Yugoslavia
IHRL International Human Rights Law
ILO International Labour Organisation
MoFA Eritrean Ministry of Foreign Affairs
NGO Non Government Organisations
NCEW National Confederation of Eritrean Workers
NUEW National Union of Eritrean Women
NUEYS National Union of Eritrean Youth and Students
OCHA UN Office for the Coordination of Humanitarian Affairs
OHCHR UN Office of the High Commissioner for Human Rights
OGNs Operational Guidance Notes
PFDJ People’s Front for Democracy and Justice
PM People’s militia
SLM Swiss State Secretariat
UKFFM UK Fact-Finding Mission
UNCOI UN Commission of Inquiry
UNHCR UN High Commissioner for Refugees
USSD US State Department
WYDC Warsai Yikealo Development Campaign
DECISION AND REASONS
INTRODUCTION
1.
A
new country guidance case on Eritrea
confronts greater challenges than usual
because of the fact that presently views about the nature and extent of the
risk awaiting Eritreans faced with forcible return to their country are
extremely polarised. On one side, there is a solid phalanx of reputable bodies
and individuals including the United Nations Commission of Inquiry (UNCOI),
UNHCR, Amnesty International (AI) and Human Rights Watch (HRW) who contend or
imply that we should maintain or extend the risk categories identified by the
Tribunal in its existing country guidance in MA
(Draft evaders – illegal departures – risk)
Eritrea
CG [2007] UKAIT 00059 and MO (illegal exit – risk on return)
Eritrea
CG
[2011] UKUT 190 (IAC). Ranged on this side are those representing the
appellants who are joined by UNHCR as intervener in arguing that the situation
in
Eritrea
has worsened. In support of this view UNHCR among others points out
that in the first 10 months of 2014, the number of Eritrean asylum seekers arriving
in Europe nearly tripled, from 13,000 the previous year to 37,000. In 2014
Eritreans were the second largest group after Syrians apprehended at European
Union external borders trying to enter in an irregular manner and the second
largest group of asylum seekers in the European Union. Many of the fatalities
in the Mediterranean were said to be Eritrean. They consider that this increase
reflects worsening conditions in
Eritrea
.
2.
On
the other side of the divide, there are mainly government bodies concerned with
Country of Origin Information (COI) and certain academics and journalists. They
do not dispute that there continues to be deep concerns about the human rights
situation in Eritrea
, but maintain that positive changes have taken place there
which entail that the views of the abovementioned bodies greatly exaggerate the
risk on return for ordinary Eritreans. One such academic, Dr Tanja Müller, has
criticised what she terms “the one dimensional interpretation of
Eritrea
by a powerful human rights lobby that seeks to monopolise what the world should know about
Eritrea
– and to morally condemn those who do not fall in line”. In an article posted on 7
December 2014, she takes issue with this interpretation for portraying
Eritrea
:
“as a
dictatorship where it is simply impossible to live a normal life in any way and
where therefore people flee and endure horrific abuses while on the way—either
during their clandestine crossing of the border, or once out by human
traffickers ultimately related to the long arm of the Eritrean state. This
narrative, advanced by organisations like Amnesty International or Human Rights
Watch, is not being recognised by anybody who actually visits Eritrea
or for
example volunteers to teach at one of its colleges as a young academic, based
at a prestigious UK university has recently done for three months.”
3.
Compounding
the strong disagreements over what is actually the present situation in Eritrea
, there are strong disagreements over the methodology of sources relied on by all
three parties. Doubtless what makes this additional dimension of disagreement
so acute is that
Eritrea
has historically been reluctant to allow independent
NGOs or human rights monitoring bodies including the UN Commission of Inquiry to
operate in the country and as a result almost all of the sources about what is
going on in the country are based on information obtained indirectly, e.g. by
members of the Eritrean diaspora or academics who have networks of individuals
inside the country. This state of affairs is one of the reasons why both in MA
and MO and now in this case particular focus has been placed on the
evidence of Professor Kibreab (PK). One of the triggers for the Tribunal’s
decision to undertake fresh country guidance was the decision taken by the Danish
Immigration Service (DIS) in 2014 to try and rectify the relative lack of
direct information from inside
Eritrea
by conducting a Fact-Finding Mission to
the country. The publication in November 2014 of this mission’s report, “
Eritrea
– Drivers and Root Causes of Emigration, National Service and the Possibility
of Return: Country of Origin Information for Use in the Asylum Determining
Process” (hereafter “DFFM Report”) and the subsequent
reliance on it by, inter alia, the UK Home Office, sparked intense controversy
in which concerns about methodology have featured prominently. The appellants and
the Intervener UNHCR have voiced similar concerns about a UK Fact-Finding
Mission (hereafter “UKFFM”) carried out in February 2016.
The true meaning of country guidance
4.
Before
turning to the task of essaying new country guidance on Eritrea
, we address one general matter about the meaning of “country guidance” in the United Kingdom context.
5. During the hearing we raised with the respondent our concern about the emergence within Home Office country publications of references to “country….guidance”.
6. We drew attention to the witness statement of Mr Martin Stares dated 24 March 2016 which explained that the Country Policy and Information Team (CPIT) was formed in March 2014 by bringing together the Country Specific Litigating Team (CSLT) and the Country of Origin Information Service (COIS). The CSLT and COIS had themselves been formed in 2008 as a result of a split in the Country Information and Policy Unit (CIPU). His statement attests that:
“[o]ne of the key reasons for bringing CLST and COIS back together was to promote greater coherence between the guidance the Home Office provided to caseworkers (primarily through the use of Operational Guidance Notes (OGNs)) and the country of origin information which informed and underpinned it. On occasions, for example, due to competing priorities within the two distinct teams, the respective products on a particular country were not updated in tandem. “
7. The CPIT team’s purpose is described by Mr Stares as being “to provide relevant, reliable, up-to-date country of origin information as well as advice and guidance on handling country specific cases to support accurate, high quality, consistent and timely decision-making”. At para 20 he states that “[t]he process of obtaining country guidance information can be summarised as follows: [9 steps are then set out].”
8. We would observe, as we did in the hearing, that the concept of country guidance is a long-established part of the UK legal system and Practice Directions identify “country guidance” as an emanation of the Upper Tribunal (formerly the AIT and IAT). The country guidance cases of the Tribunal have a high profile on European and international websites reporting recent cases and case law, e.g. UNHCR’s Refworld. The European Court of Human Rights (ECtHR) - accepted as a supervisory supranational court by 47 European countries - in particular makes frequent reference to the Tribunal’s country guidance cases. It is entirely legitimate of the Home Office to issue not just Country of Origin information but also policy and operational guidance setting out the position of the UK government. The fact that the Upper Tribunal (“UT”) (unlike the Home Office) is not in a position to update its guidance on different countries regularly only underlines the need for the executive to identify its own position on a regular basis so that caseworkers can make decisions based on the latest evidence. It is most unfortunate, however, that it has now dropped the adjective “operational” (as in “Operational Guidance Note”) and paired the term “guidance” with country “information”. This new terminology runs a real risk that members of the public and overseas readers (including courts and tribunals) might mistakenly think that “CIG” is an emanation of a UK executive body, not of the UK judiciary. When for example a reader outside the UK studies what was said in para 1.3.8 of the March 2015 CIG about the existing Tribunal country guidance case of MO (“[c]onsequently, the guidance outlined in MO above should no longer be followed”) and in the September version of the same CIG (that “MO is too prescriptive about everyone being at risk and/or the exceptions appear to be wider than those listed”), he or she could scarcely be blamed if they wrongly gleaned that this was the position under the law of the United Kingdom and if they failed to appreciate that the Home Office has no legal competence to decide whether or not a UT country guidance case is to be followed or not. It is disconcerting to note in this regard that the March 2015 CIG at 1.3.4 also contains a misrepresentation of the MO guidance in that it is stated that “Eritreans who left illegally are no longer considered per se to be at risk of harm or mistreatment amounting to persecution on return”. That was never the position set out in MO: see [133]. Absent statutory instruction, the production of “country guidance” is solely a matter for the Tribunal and the courts. No adverse comment could have been made if these statements had been accompanied by the qualification that MO was no longer to be followed by caseworkers or had made clear that in the absence of a more up to date country guidance case, caseworkers were entitled to take the view that more recent evidence enabled them not to follow MO in full or in part. But bald utterances lacking qualifications of this kind court real confusion. Given that the term “country guidance” is an established term to describe judicial guidance, we deprecate any adoption of terminology that confuses this important fact.
9. Responding to our raising of this matter during the hearing, Mr Rawat said that the term used by the Home Office was not “country guidance” but “country information and guidance”. However, as can be glimpsed from our quotation above from para 20 of Mr Stares’ witness statement, this leaves open that wherever information is not the relevant issue the Home Office is referring to “country guidance”. We express our hope that consideration will be given to terminology that maintains a proper demarcation of the role of the executive and judiciary in the area of evaluation of country conditions and risk in the field of international protection.
The country guidance issues
10. At case management hearings in the last quarter of 2015 it was directed that the issues to be determined by the Upper Tribunal (UT) in these appeals were:
“(i) The extent to which MO (illegal exit-risk on return) CG [2011] UKUT 190 (IAC) and MA (draft evaders; illegal departures; risk) CG [2007] UKAIT 59 properly reflect current country conditions and/or risk on return.
(ii) The factors likely to affect
the risk faced by those returning to Eritrea
. Relevant factors (actual or
perceived, singly or in combination) might include (a) unlawful exit; (b) age;
(c) matters arising from military conscription, draft evasion or desertion
including exemption on mental health grounds; (d) returning as a failed asylum
seeker; (e) the approach of the Eritrean Authorities to the assessment of
mental health difficulties for the purpose of exemption for someone who is
otherwise eligible for National Service.
(iii) The evidence required to support a claim and the circumstances in which inferences might be drawn.
(iv) The weight to be attached to the background material and, in particular, the Danish fact-finding mission reports and the evidence of Professor Kibreab [PK].
(v) Whether in the cases of MST and TM, the First-tier Tribunal Judge made an error on a point of law. (It will be for the Tribunal to decide how the error on a point of law is to be determined.)”
11.
As
regards (iii) and (iv) above, they are matters which were covered extensively
in submissions and will be addressed specifically later, but we do not regard
them as being country guidance issues stricto sensu, since deciding what
the relevant evidence is and what weight to attach to background country
material are rather necessary preludes to deciding such issues; they are not
the issues themselves. Further, as we shall expand on below, in some respects
they raise legal rather than country guidance issues. With the benefit of
hindsight, it is now clear that, pertinent as the DFFM Report and PK’s evidence
remain, they were an incomplete snapshot even of key items of evidence at that
time; and since then there have been others, for example, the AI Report of
December 2015, “Just Deserters: Why Indefinite National
Service in Eritrea
has created a Generation of Refugees” (hereafter “Just
Deserters” Report) and the two UNCOI Reports of 2015 and 2016,
reports which merit specific consideration just as much.
12.
In
submissions regarding (iv) above, the respondent proposed that, rather than
being excised, it be modified to read “[t]he approach to the assessment of
background source material on Eritrea
”. Whilst the clashes between the parties
over the issue of sources and methodology are important enough to be dealt with
in a separate section below, we remain of the view that to identify it as a
country guidance issue as such would shift proper focus away from findings on
country conditions to the methodology underlying such findings. Findings on the
latter are called for and specific findings will be made on the DFFM report and
PK’s evidence (and other key items of evidence); but we will no longer include
them in our list of country guidance issues.
13. In relation to (v), the UT on 24 March 2016 found that in the case of MST the First-tier Tribunal had erred in law and so his case came before us on the same basis as AA and MYK, namely (as will be identified in a moment) for a decision to be re-made on their appeals. As regards TM, it was decided by the UT on 24 March 2016 that there was no error of law and that his case would be severed from the country guidance cases.
14. In the appellants’ and UNHCR’s skeleton arguments a new country guidance issue was proposed, namely whether the Eritrean system of military service amounted to slavery, servitude or forced or compulsory labour contrary to Article 4 ECHR. In proposing this as a further issue, these parties highlighted relevant findings on it made by the two recent UNCOI Reports. Given that it was an issue thrown up by recent evidence in the case and that the respondent has had ample opportunity to address it in subsequent rejoinders, we treat the above list of issues (so far reduced to (i), (ii)) as being supplemented by a new (iii) as follows:
“(iii) Whether the Eritrean system of military service gives rise to a real risk on return of exposure to treatment contrary to Article 4 ECHR.”
Existing country guidance
15.
The
most recent country guidance case on Eritrea
is the 2011 UT decision in MO. This reaffirmed with some
modifications the 2009 country guidance decision of MA. MA in turn supplemented and amended IN
(Draft evaders – evidence of risk)
Eritrea
CG [2005] UKIAT 00106, KA
(draft-related risk categories updated)
Eritrea
CG [2005] UKAIT 00165, AH
(Failed asylum seekers – involuntary returns)
Eritrea
CG [2006] UKAIT 00078
and WA (Draft-related risks updated – Muslim Women)
Eritrea
CG [2006] UKAIT 00079.
16. In the headnote to MA, it was stated:
“1.
A person who is reasonably likely to have left Eritrea
illegally will in
general be at real risk on return if he or she is of draft age, even if the
evidence shows that he or she has completed Active National Service,
(consisting of 6 months in a training
centre and 12 months military
service). By leaving illegally while still subject to National Service,
(which liability in general continues until the person ceases to be of draft
age), that person is reasonably likely to be regarded
by the authorities of
Eritrea
as a deserter and subjected to punishment which
is persecutory and amounts to serious
harm and ill-treatment.
2. Illegal
exit continues to be a key factor
in assessing risk on return.
A person who fails to show that he or she left Eritrea
illegally will not in
general be at real risk, even if of draft age and whether or not the
authorities are aware that he or she has unsuccessfully claimed asylum in the United Kingdom.”
17.
In GM (Eritrea
) v Secretary of State for
the Home Department [2008] EWCA Civ 833 the Court of Appeal upheld the approach in MA
that illegal exit by an Eritrean applicant of or approaching draft age who was not medically
unfit, could not be assumed
where that person had been
found wholly incredible. In relation to one of the appellants, MY, who was a
17 year old girl, Laws LJ, with whom Dyson LJ agreed, said this at [53]-[55]:
“53.
… The fact (if it be so) that it is reasonably likely that any 17 year old girl
from Eritrea
, about whom nothing else relevant is known, left the country
illegally does not entail the conclusion that this particular 17 year old girl
did so. The reason is that the probability that a particular person has or has
not left illegally must depend on the particular facts of her case. Those facts
may produce a conclusion quite different from that relating to illegal exit by
members of such a class of persons about whose particular circumstances,
however, the court knows nothing more than their membership of the class. There
may indeed be a general probability of illegal exit by members of the class;
but the particular facts may make all the difference …
54. The position would only be otherwise if the general evidence was so solid as to admit of only fanciful exceptions; if the court or tribunal concluded that the 17 year old must have left illegally whatever the particular facts.
55.
Is that the position here? I do not think that it is. The categories of persons
found by the AIT in MA (largely founded on Dr Kibreab’s [PK] evidence)
to be candidates, or promising candidates, for exit visas, were not held to be
closed or watertight … It is also notable that the AIT’s conclusion
about the chances of a young male obtaining a visa is
expressed (paragraph 357) in terms of
unlikelihood only. Moreover I read paragraph 449, cited by Buxton LJ at
paragraph 13, as showing that the AIT in MA itself considered proof of
an appellant’s particular circumstances to be an important factor in
determining whether the appellant left Eritrea
illegally.”
18.
In
MO, the UT explained at [3-4] that except for one point
of clarification it did not seek to re-examine the
guidance given by the Tribunal in MA on the issues of the nature of
military and national service in Eritrea
, demobilisation and risk on return to persons who are or would be perceived
as draft evaders or deserters.
19. In MO at [133 iv] the UT modified the guidance given in MA in respect of limited exceptions who it considered would not be at risk on return as follows:
“(iv) The general position
adopted in MA, that a person
of or approaching draft age (i.e. aged 8 or over and still not above
the upper age limits for military service, being under 54 for
men and under 47 for women) and not medically unfit who is accepted as having
left Eritrea
illegally is reasonably likely to be regarded with serious hostility
on return, is reconfirmed, subject to limited exceptions in respect
of (1) persons whom the regime’s military and political leadership perceives as
having given them valuable service
(either in
Eritrea
or abroad); (2) persons who are trusted family members of,
or are themselves part of, the regime’s military or political leadership. A
further possible exception, requiring
a more case-specific analysis, is (3) persons (and their children born
afterwards) who fled (what later became the territory of)
Eritrea
during the war of independence.”
20. On the issue of illegal exit, the UT in MO at [133 iii] modified the guidance given in MA as follows:
“(iii)
The general position concerning illegal exit remains as expressed in MA,
namely that illegal exit by a person of or approaching draft age and not
medically unfit cannot be assumed if they had been found wholly incredible.
However, if such a person is found to have left Eritrea
on or after
August/September 2008, it may be, that inferences can be drawn from their
health history or level of education or their skills profile as to whether
legal exit on their part was feasible, provided that such inferences can be
drawn in the light of the adverse credibility findings.”
21. The reference to August/September 2008 arose because of the Tribunal’s finding in MO that at this time there was credible evidence of the Eritrean authorities suspending exit visa facilities albeit the UTIAC recorded that the facility had been re-opened “on a more limited basis” ([113-114]).
22. The UT in MO at [133(v)] also nuanced the guidance given in MA regarding failed asylum seekers as follows:
“(v) Whilst it also remains the position that failed asylum seekers as such are not generally at real risk of persecution or serious harm on return, on present evidence the great majority of such persons are likely to be perceived as having left illegally and this fact, save for very limited exceptions, will mean that on return they face a real risk of persecution or serious harm.”
Strasbourg
cases on Eritrea
23.
In
many country guidance cases the Tribunal attaches singular importance to lead
cases of the ECtHR dealing with country conditions and risk categories in the
relevant country. In the case of Eritrea
, however, the searches undertaken by
the parties (confirming our own understanding) revealed that there have been
very few such cases. It may be this is a reflection of the fact that in the
past decade very few forcible returns have been undertaken by European
countries. None of the cases identified are recent, which in itself, reduces
their potential value for us as source of evidence and judicial evaluation.
Decisions of national
courts and tribunals on Eritrea
24. Our attention has not been drawn to any national cases decided by courts and tribunals applying much the same EU asylum law as do we; this again may be because of the policy adopted by most EU governments of not enforcing returns.
The legal issues
25. In the course of submissions it became apparent that there was a dispute between the parties over several legal issues, namely the significance of UNHCR guidelines; the significance of previous country guidance; the role of expert evidence; methodology of sources (including anonymity of sources); and the proper test for deciding risk on return contrary to Article 4 ECHR which prohibits slavery, servitude and forced labour. In the event we decided we could resolve all but the last by reference to established case law.
The appellants
26. As the Tribunal has said on many occasions, every country guidance case has two dimensions, general and individual. The general one, which is what defines it as such, is the assessment of country conditions in relation to risk on return. But the vehicle for such assessment is always the individual appeal or appeals and in this case we are tasked with deciding the three appeals of MST, MYK and AA. Their evidence, of course, informs the wider body of evidence we have to consider regarding country conditions, but it is convenient for us to set out our conclusions upon their particular cases separately, in the last part of our decision.
27. The evidence of the appellants is set out in some detail in Appendix I. This includes their written evidence and, in the case of MST and MYK, their oral evidence before us (AA did not give oral evidence). The bare elements of their claims can be summarised as follows:
1). MST,
who was aged 27 at the hearing before us, submits that he is at risk on return
because he left Eritrea
illegally and will be viewed as a deserter. He will be
forced back into national service on return.
2). MYK,
who was aged 31 at the hearing before us, submits that he is at risk on return
because he left Eritrea
illegally and will be viewed as a deserter. He will be
forced back into national service on return.
3). AA, who was aged 29 at the hearing before us, is a paranoid schizophrenic. He submits that he will be forced to undergo national service on return notwithstanding his mental health.
Procedural matters
28.
The
case management stages of this case brooked a number of difficulties which
required treatment by a series of Directions and an Interlocutory decision now
reported as MST and Others (disclosure- restrictions- implied undertakings
Eritrea
) [2016] UKUT 337 (IAC). It is not necessary for us to deal with
these again in this decision and we have decided that in assessing the evidence
of PK we will not treat as adverse to him his failure to comply with agreed
deadlines. We would however underline that in view of the difficulties
encountered during the CMR stages of this case, the President of the UTIAC will
be reviewing whether it is necessary to amend existing Presidential guidance to
include an express warning about the possibility of an order being made against
representatives for wasted costs in the event of default arising out of
avoidable delays on the part of an instructed expert. As this case amply
demonstrates, country guidance cases can involve a great deal of preparation,
industry and effort from the parties and a considerable investment of time on
the part of the Tribunal. They can encompass (as does this case) judicial
decision-making likely to have very significant implications for the processing
of many asylum claims in the UK and beyond. It cannot be allowed that
scheduling of such cases is sabotaged by a lack of due diligence on the part of
experts.
29. Although a limited number of other procedural matters arose in the course of the hearing before us, it is not necessary to say anything about them in this decision, there being agreement between the parties as to their resolution.
A. EVIDENCE AND SOURCES
30. We have chosen to structure this decision so that we identify and discuss the most relevant evidence at the same time as we seek to decide key issues, a task which has been made much easier for us by the careful and detailed submissions made by the parties regarding relevant background evidence and sources. It will help lay the foundations for our ensuing assessment, however, if we briefly identify the main reports and sources identified to us. For this purpose we shall adopt a chronological order, except where there are closely-related reports or items which are most conveniently dealt with together. The Home Office and Foreign and Commonwealth Office materials are dealt with separately towards the end of this section since they include up-datings and have been the specific focus of the appellants’ submissions arguing that the Home Office was wrong to issue policy statements saying that MA and MO were no longer to be followed in certain respects by caseworkers.
1. Background evidence
UNHCR
Eligibility Guidelines for Assessing the International Protection Needs for
Asylum-seekers from Eritrea
, 20 April 2011
31.
In
MO, the Tribunal had before it UNHCR Eligibility Guidelines on Eritrea
dated April 2009. In April 2011 UNHCR issued updated Guidelines which remain in
place. Its list of “risk profiles” reads as follows:
“UNHCR considers that individuals with the profiles outlined below require a particularly careful examination of possible risks. These risk profiles, while not necessarily exhaustive, include (i) persons avoiding military/national service; (ii) members of political opposition groups and Government critics; (iii) journalists and other media professionals; (iv) trade unionists and labour rights activists; (v) members of minority religious groups; (vi) women and children with specific profiles; (vii) lesbian, gay, bisexual, transgender and intersex (LGBTI) individuals; (viii) members of certain ethnic minority groups; and (x) victims of trafficking.”
Eritrean Ministry of Information publication: “UNHCR Eligibility Guidelines: Factual Findings or Recycled Reformation”, 17 December 2015
32.
Although
over four years later, it is convenient to note this publication emanating from
the Eritrean government attacking the 2009 and 2011 UNHCR Guidelines for
“sloppy, cut-and-paste desk research” characterised, it was said, by “wholesale
regurgitation of prevalent, negative literature on Eritrea
from biased and
politically motivated entities”.
33. The report goes on to deny that forcible returns from Libya and Egypt had resulted in any ill-treatment, asserting that the government in fact exercised clemency to them. In response to the UNHCR statement that “[f]or some Eritreans, being outside the country may be sufficient cause on return to be subjected to scrutiny, reprisals and harsh treatment”, this publication states:
“No Eritrean is subjected to
harassment simply because he/she lives abroad. The fact is even those who have
asylum papers come back to their country periodically for family reunion,
vacation and other personal matters. Eritrea
’s tourism is largely based on the
Diaspora who visit their country in summer as well as during the Christmas,
Easter and Independence Day celebrations. More than 85,000 Eritreans come back
for vacation every year and this number is greater on special occasions, as
will be the case in 2016 when Eritreans will celebrate next May its
Independence Silver Jubilee....”
34.
UNHCR
has confirmed in writing to this Tribunal that UNHCR’s position remains as set
out in the 2011 Guidelines, stating that before updating them UNHCR would
ideally wish to have full information based on full access to the country.
UNHCR also confirmed that in collaboration with the government of Eritrea
it extends assisted protection and assistance to an average of 2,450 Somali
refugees based in a camp in Massawa and also has open technical collaboration
with the government on issues of mixed migration that have an asylum nexus.
Danish Fact-Finding Mission (DFFM) Reports, 25 November and 16 December 2014
35.
As
already noted, on 25 November 2014, the DIS published its DFFM Report. This report
was said to be based on three 2014 fact-finding visits by Danish officials to London, Ethiopia and Eritrea
. The report comprises a 20 page report (which where
appropriate we shall call its “main body”) and 54 pages of annexed
interview/meeting transcripts.
36.
The
main body of the report is organised into five parts, part 4 dealing with
National Service and part 5 with Return to Eritrea
. In 4.8 under the heading
‘Consequences for evasion/desertion from the national service’, it is said
that:
“... [PK], ......, also stated that
over the past two or three years, the government’s attitude towards NS [national
service] seems to be more relaxed. It is now possible for evaders and
deserters who have left Eritrea
illegally to return if they pay the two per
cent tax and sign the apology letter at an Eritrean Embassy. Finally, [PK] was
aware of a few deserters from the national service who have visited
Eritrea
and safely left the country again.”
37.
The
report explains that in view of the fact that existing reports on Eritrea
were to a large extent based on sources outside the country, it was decided there
was a need for updated and first-hand description of the conditions on the
ground. It states that in order to prepare and plan the missions to
Eritrea
and Ethiopia, the DIS conferred with other immigration authorities in Europe as well as with
PK. It states that in
Eritrea
and Ethiopia the delegation consulted
representatives of Western embassies, UN agencies, international organisations,
international non-governmental organisations, local non-governmental
organisations, a well-known Eritrean intellectual as well as a representative
from the Eritrean government. To the best of its knowledge, avows the main
body of the report, the consulted interlocutors represented “a broad spectrum
of competent sources knowledgeable on the relevant issues in
Eritrea
”.
38. On the same day the DIS announced that on account of this report it was changing its policy so that “illegal exit will not in itself amount to persecution or grant the right to international protection”.
39. In early December 2014, PK publicly disassociated himself from the report. He said to the Danish newspaper Politiken that “I felt betrayed, I demand my name be taken off the report”. He said that his views had been misrepresented and that the Danish authorities had “basically ignored a lot of facts and hand-picked a few that fit the conclusion”.
40.
On
2 December, AI condemned the report as “completely absurd”. David Bozzini, a
renowned expert on Eritrea
, publicly supported PK’s criticisms. On 9 December
2014, two of the report’s researchers, Mr Olsen and Mr Olesen, publicly
criticised the report. (We should mention at this point that the evidence in
this case includes an unsigned 6-page “Statement on Danish
Eritrea
Report” from
these two persons, dated 28 April 2016, which gives their account of their
involvement in the report and its aftermath.)
41. On the same day the DIS ostensibly backtracked, announcing that the Danish authorities would continue to recognise as refugees Eritreans fearing persecution as a result of their illegal exit and/or desertion or draft evasion and that “this might well involve providing the benefit of the doubt to the asylum-seeker” and that it expected to recognise such asylum claims in “many cases”.
42. On 16 December 2014, the DIS published a new Appendix Edition of its 25 November report leaving in all of PK’s contributions but showing them as crossed out.
43.
The
above summary does not mention all the organisations and individuals who came
forward to voice criticisms of the DFFM Report. Their number included UNHCR who
in December issued a three page Note, welcoming the decision of the DIS to
produce a COI Report but expressing a number of concerns as regards the
methodology used in the report. It stated that the main body of the report made
selective use of information provided by interlocutors, including statements
that could not be traced to these interlocutors’ statements as reviewed and
cleared by them. It objected to the lack of proper understanding of the
regulatory framework for the media, NGOs and other actors in Eritrea
. In later materials, UNHCR has criticised the DFFM Report for treating its sources in Asmara as a “multitude of independent sources” rather than as one source with the
additional understanding that members of the international community in Asmara have limited freedom of movement. On 17 December HRW published an analysis entitled
“Denmark:
Eritrea
Immigration Report Deeply Flawed”. The Danish press media
carried many articles on the controversy that had engulfed the report and the
responses of various governmental and political actors in Denmark.
44.
On
16 September 2015 the Danish Parliamentary Ombudsman published a report on the
Eritrean case. His report makes clear that he was given access to all relevant
information held by the DIS and the government. Whilst criticising the public
pronouncements made by government sources about the DFFM Report and finding
that the general public had been left with a very unclear view of what the
basis was for the completely conflicting statements from the DIS on 25 November
and 9 December, he found no breach of the applicable law. His report mentions
that Mr Olsen and Mr Olesen had referred the matter to him on 27 January 2015
asking him to “go into the case regarding the employment law warnings they had
received and the facts and circumstances around what became of the fact-finding
report on Eritrea
”. Despite further noting that the personnel matters
concerning these two had been settled (upon the service warnings given to them
having been revoked), his investigation addressed a number of issues, including
“was pressure put on staff of the DIS to paint a favourable picture of
conditions in
Eritrea
which were not actually how things were?”. He concluded
that the decision to set up the DFFM was both objective and lawful and that it
“strived to ensure as broad a composition of sources as was possible”. He
declared that “I have no reason to believe that the DIS wished to give the
conclusions in the report an untenable expression or put pressure on its staff
with this purpose in mind”. Equally he was satisfied that “serious doubts have
been raised as to how the authorities deal with the case, including questions
about improper political intervention in dealing with asylum cases”. On 25
November he confirmed that he stood by his report.
Landinfo, 23 March 2015 and 16 April 2015 and May 2016
45.
On
23 March 2015, Landinfo, the
Norwegian Country of Origin Information Centre, which is an independent body
within the Norwegian Immigration Authorities, published
two reports, one entitled “Eritrea
: National Service”, and the other entitled “Response:
Reactions towards returning asylum seekers”. The National Service Report
explained that Landinfo had made four trips to
Eritrea
in the past four years,
the last completed in January 2015. This report emphasised that because of the
difficulties in obtaining information from sources inside
Eritrea
it had to rely on Eritreans outside their country. The report noted that there was no
evidence as yet that the Eritrean government had implemented its [2014] promise
to limit national service to eighteen months. It stated that the upper age
limit for conscription to national service had increased since the border war;
however, women were increasingly exempt because of marriage, giving birth or on
a religious basis. Eritreans who evade national service were said to be
probably exposed to arbitrary punishments from local commanders, and there had
been indications that Eritreans performing their national service in military
units have been more subject to punishment than Eritreans in the civilian
sector. In the “Reactions towards returning asylum seekers” document, Landinfo
addressed the claim made by AI among others that the act of submitting an
asylum application as such will lead to adverse treatment on return because
such persons are seen as traitors. It noted that there was very little certain
and verifiable information regarding this. It noted that PK had said in 2012
that he did not have specific examples of what has happened to returned asylum
seekers. After reviewing the known evidence, it concluded that “[w]e do not
currently have an empirical basis for saying that an application for asylum as
such will lead to reactions from the Eritrean authorities”.
46.
On
15 April 2015 Landinfo issued a “Response: Eritrea
: Exit visas and illegal
exit”. It stated that the vast majority of those who leave
Eritrea
do it illegally. The categories of those who could obtain exit visas were limited and there
can be additional difficulties for close family members of people who have left
the country illegally and have been critical of the government or lack
documentation. Landinfo stated that its “impression” was that the authorities
assess Eritreans returning home based on circumstances such as those
surrounding their departure, national service status, any political activity in
exile, their network in
Eritrea
and the payment of the diaspora tax. “It is
probably the reasons behind the departure that can lead to reprisals on
returning home and not the illegal departure in itself.” It considered that
persons who had restored their relationship with the authorities by signing the
retraction [repentance] letter, paying the 2 per cent tax in exile and who do
not participate in activities critical of the government were likely to be less
vulnerable to reprisals from the authorities. A good network and contacts in
the government apparatus and the party were also probably useful.
47.
We
should perhaps note here that in the Review of UK Home Office Country
Information and Guidance – “Eritrea
: National (incl. Military) Service”
(version 2.0e, September 2015) and “
Eritrea
: Illegal Exit” (version 2.0e,
September 2015), the review prepared by Dr John Campbell criticises not only
the DFFM Report but the (pre-2016) Landinfo reports also.
48.
The
new versions of the Home Office CIGs published on 4 August 2016 include a
number of references to a Landinfo thematic report on Eritrea
dated 20 May 2016. We have taken account of the contents of those references.
UN Commission of Inquiry Reports
(A/HRC/29/CRP.1), 5 June 2015 (2015 UNCOI Report)
49.
In
this Report the Commission explains that it had been set up by the Human Rights
Council under resolution 26/24. It describes Eritrea
as a country characterised
by human rights abuses, some of which may amount to crimes against humanity.
At [26] of the summary report it states that “Eritreans are fleeing severe
human rights violations and are in need of international protection”. At [46]
of the same summary report it states that Eritreans who attempt to leave the
country are seen as traitors.
50.
As
regards its methodology, the Commission states that although unable to visit Eritrea
it obtained first-hand testimony by conducting 550 confidential interviews with
witnesses residing in third countries. It also received 160 written
submissions. At [34] of the detailed report it states that in order to
establish the facts and circumstances of alleged violations and taking into
account the impossibility to access
Eritrea
, it decided to collect first-hand
testimonies and accounts of victims and witnesses of alleged human rights
violations “from Eritrean refugees, asylum seekers, migrants and other members
of the Diaspora”.
51. On the issues related to national service, the three main conclusions of the detailed report were:
“1. [T]reatment of apprehended draft evaders and deserters during detention often amounts to torture, cruel, inhumane or degrading punishment ([1389]).
2. People attempting to leave – or who have previously left – the country illegally are regarded as “serious offenders but also as traitors” ([431]) and “with a few exceptions [are] arrested, detained and subjected to ill-treatment and torture” ([444]).
3. That conditions of national service are characterised by lack of adequate food, access to water, access to hygiene facilities and adequate accommodation during military training and service, such conditions constituting cruel, inhuman or degrading treatment ([1391]).”
52. At [61] of the summary report the UNCOI writes that:-
“National service as implemented by the Eritrean authorities involves the systematic violation of an array of human rights on a scope and scale seldom witnessed elsewhere in the world”.
53. At [1397] of the detailed report it states that:
“The indefinite duration of national service: its terrible conditions and treatment including arbitrary detention, torture, sexual and gender-based violence, forced labour, absence of leave and the ludicrous pay; the implications this has on the possibility of any individual to form a family, have a family life and to have favourable conditions of work, make national service an institution where slavery-like practises occur.”
(A/HRC/32/CRP.1), 8 June 2016 (2016 UNCOI Report)
54.
In
June 2016 the Commission published a further report, releasing its detailed
findings on 8 June. It noted that its further report arose as a result of the
Human Rights Council in its resolution 29/18 having extended its mandate for
one year “to investigate systematic, widespread and gross violations of human
rights in Eritrea
with a view to ensuring full accountability, including where
these violations may amount to crimes against humanity”. Its two principal
conclusions were first that during the period under review there had been no
improvement with respect to the most critical human rights violations in
Eritrea
documented in its first report; and second that the Commission had
reasonable grounds to believe that crimes against humanity, namely enslavement,
imprisonment, enforced disappearance, torture, other inhuman acts, persecution,
rape and murder, have been committed in
Eritrea
since 1991.
55.
The
2016 UNCOI Report notes that in response to its call for responses it received
almost 45,000 written submissions, “the vast majority of which were critical of
the first report of the Commission”. In a communication of 20 June responding
to a question the Tribunal had raised during the hearing, Ms Dubinsky relayed
confirmation from the Commission that for its 2016 report “the Commission of
Inquiry has interviewed 123 witnesses since the issuance of [its] first report
in June 2015, many of them individuals who left Eritrea
in the period
2014-2016”. In response to further directions addressed to UNHCR, the UT
received a letter dated 15 August 2016 from the Special Rapporteur on the
situation of human rights in
Eritrea
and who was a member of the Commission,
explaining why the Commission had concluded that the information provided in
the 44,267 written submissions did not have a bearing on the information
provided by “the more than 833 other sources of information”.
56. Observing that the “campaign critical of its first report was well organised”, that most critics had not read the report and appeared to rely on erroneous understandings or deliberate misinformation and that it had evidence that some letters had been submitted involuntarily, the Commission concluded that the submissions did not undermine the findings described in its first report.
EASO Country of Origin
Information Report, Eritrea
Country Focus, May 2015
57.
This
Report highlights the difficulties of access to relevant COI about Eritrea
, which has led to reports on sensitive issues having to rely largely on sources outside
Eritrea
. It notes that the few available reports based on research in
Eritrea
mainly drew on government statements and anecdotal knowledge of international
representatives, and not on first-hand information. “This difficulty was
demonstrated in recent polemics regarding the Danish fact-finding report”. The EASO
Report describes itself as being based on publicly available reports of COI
units, UN agencies, human rights organisations, solicitors, officials, NGO
papers, government and diaspora media. It states that it has been completed
with information obtained from interviews, e.g. during information gathering
missions. We shall refer later to what this report has to say about certain
topics, that of lawful exit categories in particular. We mention here that at
3.8.2, in a section headed “Punishment for returning deserters and draft
evaders”, the EASO Report states that there have been no new empirical findings
since 2008 and therefore the punishment currently imposed on deserters and
draft evaders is difficult to establish. “However, most sources state that
punishment is imposed arbitrarily on an extra-judicial basis without regard for
the laws”. It notes that there have been many instances of overland
repatriations from Sudan in recent years but that there is no information
available on the fate of those repatriated after their return. Reference is
made to a HRW Report ’Sudan: Stop Deporting Eritreans’, 8 May 2014 and a UN
News Centre, ‘UN refugee agency warns Sudan over forced return of Eritrean
asylum seekers’, 4 July 2014. The Report adds:
“Some of the respondents contacted
in Eritrea
during Denmark’s and Norway’s fact finding missions in late 2014 and
early 2015 believed that deserters and draft evaders were held in prison for
several weeks or months and were then reassigned to NS [national service].
However, several of the experts consulted in 2013 and 2014 by Norway, the
Netherlands and Denmark believed that repatriated deserters and draft evaders
may still be subjected to interrogation and mistreatment…The Eritrean
leadership has stated on several occasions that those returning to the country
will not be punished as long as they have not committed offences but it has not
yet been made clear whether desertion, draft evasion or illegal exist are
regarded as offences….”.
58. At 6.4.4 the Report states that “[t]he Eritrean authorities claim that people who have left the country illegally may return without fear of punishment after they have paid the diaspora tax and signed the repentance form, but they may be sent to a six-week training course ‘to enforce their patriotic feelings’”.
US State Department
Country Reports on Human Rights Practices, chapter on Eritrea
, 2014 and 2015
Chapter on Eritrea
2014
(25 June 2015)
59. The USSD Report covering 2014 stated that refusal to perform military or militia service, failure to enlist, fraudulent evasion of military service, and desertion were punished by lengthy imprisonment or other harsh arbitrary forms of punishment. The report stated that “the government did not demobilise many conscripts from the military as scheduled and forced some to serve indefinitely under threats of detention, torture, or punishment of their families”.
60. As regards exit visas, the same report stated that during the year:
“the government imposed new restrictions. Authorities generally did not give exit visas to children ages 5 or older. In September members of the civilian militia were told that any men or unmarried women in the civilian militia would be unable to get an exit visa until further notice. Categories of persons most commonly denied exit visas included men under age 54, regardless of whether they had completed the military portion of NS [national service] and women younger than age 47. The government did not generally grant exit permits to members of the citizens militia, although some whom authorities demobilised from national service or who had permission from their zone comrades were able to obtain them. Authorities arrested persons who tried to cross the border and leave without exit visas. A shoot-to-kill policy was in effect for those attempting to cross the border to exit the country without authorisation.
To prevent emigration, the government generally did not grant exit visas to entire families or both parents of children simultaneously. Some parents avoided seeking exit permits to children approaching the age of eligibility for national service due to concern that they would be denied permission to travel, although other adolescents were granted exit permits. In the past diaspora males who visited the country reported being required to pay a two per cent tax on foreign earned income before being given exit visas. This was not commonly enforced.”
Chapter on Eritrea
2015
(13 April 2016)
61. The USSD Report covering 2015 is in similar terms, with some updating, including numerous references to the 2015 UNCOI Report. In section 7 it states:
“Forced labour occurred. Despite the 18 month limit on national service under the law, the government did not demobilise many conscripts from the military as scheduled and forced some to serve indefinitely under threats of detention, torture, or punishment of their families, persons performing national service could not resign or take other employment, generally received no promotions or salary increases, and could rarely leave the country legally because they were denied passports and/or exit visas. Those conscripted into the national service performed standard patrols and border monitoring in addition to labour such as agricultural, terracing, construction and laying power lines. In its examination during the year of forced labour in the country, the ILO Conference Committee on the Application of Standards noted discussion “relating to the large-scale and systematic practice of imposing compulsory labour on the population for an indefinite period within the framework of the national service program which encompasses all areas of civilian life and was therefore much broader than military service.”
62. The report also deals with foreign travel and exit visas and the following is stated; “Authorities generally did not give exit visas to children ages five and older. Some parents avoided seeking exit permits for children approaching the age of eligibility for national service due to concern they themselves would be denied permission to travel, although some adolescents were granted exit permits. Categories of persons most commonly denied exit visas included men under age 54, regardless of whether they had completed the military portion of national service, and women younger than 30, unless they had children.”
Amnesty International, Report on AA, 22 September 2015, and the “Just Deserters” Report, December 2015
Report on AA, 22 September 2015
63. AI produced a report for the case of AA dated 22 September 2015.
64. In addition to dealing with the particular case of AA, this report expressed its concern that the present Home Office CIGs [see below] continued to be flawed particularly because they continued to rely on the DFFM Report as their primary justification for their guidance positions. The report contains a specific critique of various aspects of the DFFM Report, particularly its reliance on two sources, a regional NGO based in Asmara and a well-known Eritrean intellectual, both of whom it said were likely to be pro-government. It considered that diplomatic sources in Asmara were “highly likely to be prevented from obtaining relevant information and therefore risk speaking beyond what their actual evidence above would support.” It stated its view that Eritrean national service amounts to a system of forced labour in its own right.
65. In this report AI also takes issue with what it considers to be the implication that “there is something inherently unreliable about human rights information sourced from nationals external to their country who may or may not ultimately be seeking, or who have previously sought, international protection in a western country”. This report emphasises that if there are reasons to be sceptical about the veracity of a source’s account, these are considered and investigated to the greatest possible extent before a decision is made about whether or not to rely on the information being provided.
66. The report then explains and summarises the recent research it did which was subsequently published as “Just Deserters” (see below).
67. The remaining parts of the report are devoted to assessment of medical fitness for national service conscription and availability of mental health services.
“Just Deserters” Report, December 2015
68.
In
December 2015, AI published its study “Just Deserters: Why Indefinite National
Service in Eritrea
has created a Generation of Refugees”.
69.
The
report stated that AI conducted face-to-face interviews with 72 Eritreans who
had fled from Eritrea
between July 2014 and July 2015. Corroborating
information was also taken from a further fifteen interviews with Eritreans who
had left
Eritrea
illegally in 2013 and 2014.
70.
This
report repudiates what it describes as the attempt by “the authorities in
several countries where Eritreans have claimed asylum” to refute the notion
that those who flee national service have valid grounds for claiming
international protection. Contrary to the view of these authorities that there
had been an improvement in the experience of national service conscripts and
other Eritreans, AI said it found “no discernible changes in national service
practices as of November 2015”. In AI’s view national service remains the key
factor causing people to flee Eritrea
and a high number of people who leave are
unaccompanied minors:-
“Children are walking alone, often without telling their parents to another country, to avoid a life of perpetual forced labour on low pay with no genuine education or viable work opportunities through which they or their families could live.”
71. The categories of those required to do national service is said by this report to have been expanded by the introduction since 2013 of the “People’s Army”. Men as old as 67 have been re-conscripted through this system.
72. The “Just Deserters” Report observes that:
“The experiences of people caught,
arrested and arbitrarily detained for attempting to leave the country is
indicative of the likely treatment failed asylum-seekers will face if they were
forcibly returned to Eritrea
. There is a high likelihood that anyone of
approximately national service age who is returned to
Eritrea
would be subject
to arbitrary detention without charge; as is the widespread pattern, would face
possible torture or other ill-treatment to extract information on how and with
whom they left the country and then would be conscripted or returned to
indefinite national service. It is possible that some would avoid such a fate,
but as the implementation of punishment is arbitrary, the risk must be
considered to apply in every case.”
73.
The
report also highlights that the International Labour Organisation (ILO),
commenting on Eritrea
as a party to ILO Conventions, has underlined that the
large-scale and systematic practice of imposing compulsory labour on the
population within the framework of national service in
Eritrea
is incompatible
with ILO Conventions, which prohibit the use of forced and compulsory labour as
a method of mobilising and using labour for purposes of economic development.
Lifos reports on Eritrea
, 23
November and 15 December 2015
Subject report People’s
Army in Eritrea
, version 1.0, 23 November 2015
74. Lifos is the Swedish Migration Agency’s database for legal and country of origin information.
75.
Having
set out known information about the People’s Army, Lifos comments that the
People’s Army is now established throughout Eritrea
although it is likely that
it has been implemented in different degrees around the country depending on
local and regional conditions. The upper age limit seems in practice to be
around 70. Women are involved to a lesser extent. The unpopularity of the
People’s Army was demonstrated when many stayed away from the reserve training
in the autumn of 2014 but by the beginning of 2014 many people no longer dared
to refuse. It was unclear to what extent people have been punished. By the
introduction of the People’s Army the Eritrean state has in principle mobilised
the entire adult population.
Country Report Eritrea
, 15 December 2015 version 1.0
76.
This
further report by Lifos finds the human rights situation in Eritrea
to be one that “remains deeply troubling”. The most common violations are said to include
indefinite service in the national service, forced labour, torture, detention
and inhumane and degrading treatment. It is said that most sources state that
people who desert and evade the national service risk harsh penalties that may
include torture and other degrading treatment. Some information also indicates
that punishment is much milder than before. Lifos states that there seems to be
big differences between serving in the military and serving in the civil sector
for the national service, both in terms of living conditions and penalties.
Lifos concludes that it is very difficult to comment on the manner in which an
illegal exit is penalised. Most likely it is a combination of factors. “Several
sources state that asylum-seekers who return by force risk being subjected to
serious abuses, including torture…”. The report goes on to state that “from a
source-critical perspective” the information contained in the 2015 UNCOI Report
“has some weaknesses. It does not mean that it is generally not credible. It
should also be noted that some well-established sources do not always state their
sources, including the USSD, HRW and AI”. Lifos notes that there is a lack of
verifiable information about what happens to people who have had their asylum
applications rejected and are returned to
Eritrea
by force. It states that
“several sources state that asylum-seekers who are returned by force risk being
subjected to serious abuse, including torture. Some sources emphasise that the
Eritrean government is not consistent in its actions and can react in different
ways. …”
Swiss Visits, January and March 2016
77.
In
January 2016 a group of Swiss politicians conducted a private visit to Eritrea
. They were reported in the press and by the Swiss State Secretariat (SLM) to have
praised the openness of the people they met and stated they did not have
problems travelling around without surveillance. One of the group, Claud
Béglé, told a Swiss public radio that “....the system remains authoritarian but
it is opening up”. The visit was the subject of criticisms by some Swiss
politicians and the Swiss branch of AI. In a statement the SLM observed that
the politicians concerned did not discuss human rights topics related to asylum
procedures and concluded that there was “no sufficiently strong evidence to
show that the human rights situation in
Eritrea
has improved significantly”.
78.
In
March 2016 three migration officials, two Swiss and one German, conducted a
fact-finding mission to Eritrea
organised by the head of the SLM. It was
reported that they were not allowed to see prison or military facilities but,
accompanied by Eritrean officials, they visited towns and schools during a
two-week trip to gather information that could help them better understand the
situation in the country for the Eritreans who make up the largest group of
asylum seekers in Switzerland. The mission was reported to find few rights
improvements in
Eritrea
and the head of the SLM was quoted as saying that
Eritrea
had officially gone back on its word to shorten the required national military
service. He stated that “[w]e are checking whether people who go back to
Eritrea
after having left illegally could still face draconian punishments”.
79.
In
a statement to a newspaper on 9 May 2016 the head of SLM confirmed that
following the FFM in March 2016 it had been concluded that there was no
improvement in human rights and there was no indication that the duration of
national service would decrease to 18 months. The two latest Home Office
Country Information and Guidance publications on Eritrea
, both published in
August 2016, contain several translated excerpts of the report of this March
2016 Fact-finding mission (the original is in German).
Human Rights Watch reports
80.
Reference
has already been made to a December 2015 report by HRW criticising the DFFM Report
and reference will be made below in the subsection on Bisha Mines to the HRW Report,
“Hear no Evil: Forced Labour and Corporate Responsibility in Eritrea
’s Mining
Sector”, 25 January 2013.
Human Rights Watch, “Sudan: Stop Deporting Eritreans”, May 2014
81.
This
short report states that on 1 May 2014, Sudanese authorities in eastern Sudan handed 30 Eritreans over to the Eritrean security forces, according to two advocates
in close telephone contact with the group at the time. “Human Rights Watch also
obtained further credible information confirming that the deportation took
place and that six members of the group were registered refugees”. It goes on
to say that on 3 May two Eritreans from a different group told a third advocate
that a few days earlier the Sudanese authorities had intercepted a group of
about 600 persons who included Eritreans and had taken them to the town of
Donga where they were convicted and all the Eritreans were ordered to be
deported to Eritrea
.
Human Rights Watch, Human Rights Watch Report 2016
82.
The
2016 World Report highlighted “the continuing flow of Eritreans escaping the
country, and the publication of a scathing 453-page report by a UNCOI
describing the serious human rights violations prompting thousands to seek
asylum outside Eritrea
. The 2016 report also noted that the DFFM Report of
2014 had been repudiated by two of its three authors. It added:
“Despite widespread criticism of the
Danish Report, the United Kingdom’s Home Office changed its guidance about
Eritrea
in early 2015 to assert that asylum seekers “who left [
Eritrea
]
illegally are no longer considered per se to be at risk of harm or mistreatment
amounting to persecution on return”.
UK government materials
UK Home Office Policy, February 2014 – September 2015
83.
The
February 2014 Home Office Guidance Note on Eritrea
reiterated the position it
had taken in previous notes, concluding that:
1). The Eritrean government “views as political opponents those who evade military service or desert from the military” and that “the treatment of such individuals is likely to amount to persecution”;
2). Eritreans forcibly returned to their country after leaving illegally “will be subjected to arrest without charge, detention, torture and other forms of ill-treatment”; and
3). National/military service may involve abuses such as indefinite forced labour, inadequate food and medical care, arbitrary arrest and detention for minor infractions, and, in the case of women, sexual violence.
84.
In
March 2015, the Home Office published two reports, “Country Information and
Guidance, Eritrea
: Illegal Exit” and “Country Information and Guidance:
Eritrea
: National (including military) Service”. These modified its previous conclusions on
each of the three aforementioned matters as follows:
1). It was stated that “those who refuse to undertake or abscond from military/national service are not viewed as traitors or political opponents”;
2). Eritreans
fleeing national service “who left illegally are no longer considered per se to
be at risk of harm or mistreatment amounting to persecution on return”. It was
stated that anyone “who left Eritrea
illegally [is] not at risk of harm
provided they have paid income tax... and have signed a ‘letter of apology’ at
an Eritrean Embassy” before returning home.
3). As
regards conditions of national service, it was stated that “although a person
may be able to demonstrate that they would be at real risk of mistreatment or
inhuman, degrading treatment as a result of the conditions of military service,
it cannot be said that every single person undertaking some form of military
training as part of their [national service] requirement would face such
risk”. The National Service Report also concluded that “the most up-to-date
information available from inside Eritrea
suggests, in general,
military/[national service] lasts for around four years” and “[national
service] is generally between eighteen months and four years”.
Foreign
& Commonwealth Office Report, Eritrea
– Country of Concern, January 2015
85.
The
Foreign and Commonwealth Office (FCO) Report, “Eritrea
– Country of Concern”
January 2015 considered that “the Eritrean Government made no visible progress
on key human rights concerns... continued to violate its international
obligations and domestic law, including in the areas of arbitrary and inhumane
detention, indefinite [national service], and lack of ....freedom, freedom of
the media and freedom of speech. The government continued to cite ‘no war, no
peace’ with Ethiopia as justification for its failure to implement the 1997
Constitution, which provides for democratic government and fundamental rights
and freedoms”.
Country Information and
Guidance (CIG) Eritrea
: Illegal Exit, version 2.0e
86.
In
September 2015 the Home Office issued version 2.0 of its ‘Country Information
and Guidance (CIG): Eritrea
: Illegal Exit’ which stated that the categories of
persons likely to be granted an exit visa remained limited and there were large
numbers of Eritreans – reportedly thousands each month – leaving the country
illegally.
87. The CIG stated that:
“More recent information suggested that not everyone who left illegally is detained on return (or that all draft evaders are detained) and that the Eritrean authorities have neither the will nor means to imprison every returnee. The evidence suggests that whilst some are detained/imprisoned (with the length of time appearing to vary), some are fined, others are simply re-assigned to national service. If disproportionate punishment amounting to serious harm is imposed, it is applied arbitrarily”.
88.
This
CIG also sought to draw the following conclusion from the evidence that many
Eritreans returned to Eritrea
each year, for example to visit friends and
family:
“...the fact that they have e.g.
acquired foreign citizenship is not a reason, of itself, to exempt a person
from [the requirement to complete national service]. This suggests that either
those leaving Eritrea
have completed national service and/or there is no real
risk of a penalty being imposed for having previously left illegally.”
89. The CIG also attached considerable significance to evidence suggesting:
“that a person who left Eritrea
illegally, even a draft evader, can return to
Eritrea
provided they sign a ‘letter of
apology’ and pay any outstanding (2%) diaspora tax at an Eritrean Embassy. The
diaspora tax is considered a reasonable requirement and a refusal to comply
with this will mean the person is not issued with a travel document to return
to
Eritrea
voluntarily, but this would not amount to persecution or serious
harm.”
90. For these reasons the CIG concluded that “MO is too prescriptive about everyone being at risk and/or the exceptions appear to be wider than those listed”.
Country Information and Guidance (CIG) on National Service (incl. Military) Service), version 2.0
91. Also in September 2015 the Home Office issued Version 2.0 on Country Information and Guidance on National Service (incl. Military) Service.
92. This CIG accepted that the physical conditions during national service were “generally harsh” with reports of torture and mistreatment, often for very minor infractions and punishments appearing to be meted out on an arbitrary basis. However, some sources reported that conscripts were not overworked or ill-treated and that conditions vary. As regards civilian national service, conditions were generally better.
93. The CIG stated that the average time a person spends doing national service is between 4-6 years, although some spend longer: “this appears to be arbitrary”. At 3.1.9 it is stated:
“Where a person can demonstrate that there would be a flagrant denial of their right not to be required to perform forced labour, in particular beyond the 4-6 year average period of national service, then they may be entitled to a grant of DL [discretionary leave].”
94. At 3.1.10-3.1.11 it stated:
“3.1.10. The evidence suggests that while some deserters/evaders may be detained/imprisoned (with the length of time appearing to vary) some are fined, others are simply re-assigned to national service. In order for a punishment to be considered disproportionately harsh or severe, it would need to be of a particularly serious nature. Long prison sentences will not normally be enough. However, the physical conditions of detention and potential for mistreatment may be such that a person can demonstrate that they are at real risk of persecution or serious harm.
3.1.11. The risk may be higher for those who have e.g. deserted more than once and/or deserted a critical post graduate diploma in business management. However, there are other ‘critical’ posts such as teachers, which the Eritrean Government is keen to retain.”
Report by the Independent
Advisory Group on Country Information (IAGCI) on Eritrea
Country Information
and Guidance Reports produced by the UK Home Office, 13 May 2015
95. Although by an independent advisory body, and so not constituting government materials, it is convenient to include here a short summary of this report. It states that in the view of the IAGCI the two March 2015 CIG Reports were:
“marred by severe methodological concerns. In particular, where they refer to illegal exit, conditions on return and national military service, the two CIG reports rely heavily on [the DFFM Report] [which] has itself been widely criticised in terms of its methodology.”
Review of UK Home Office
Country Information and Guidance – Eritrea
, prepared for the Chief Inspector of
Borders and Immigration and the Independent Advisory Group on Country
Information (IAGCI), 15 November 2015
96. This review by Professor John Campbell was commissioned by the IAGCI and was stated as being drafted in line with instructions received through the IAGCI chair. It is focused on the two September CIGs.
97. In this report Professor Campbell argues that the September Illegal Exit CIG omitted important COI which at the very least would have qualified their policy recommendations.
98. He states that in view of the wide-ranging criticisms made of the DFFM Report the Home Office could not rely on any part of it.
Home Office Response to IAGCI Review, November 2015
99. In this document one of the principal points made by the Home Office is that the IAGCI review failed to apply the same level of objective assessment to all sources referred to in the CIG. The response states that it does not share the view that the DFFM is ‘discredited’. Professor Campbell is also criticised for simply recycling previous criticism of the DFFM Report rather than engaging with the CIG observations which only cited the agreed notes. It contends that the review had gone outside the remit of the terms of reference by reviewing the policy/guidance section.
UK Fact-Finding Mission – Visit 20 February 2016
100.
On
24 March 2016 the respondent served evidence on the Tribunal which included
“Information from the Home Office’s Fact-Finding Mission to Eritrea
(7-20 February) 2016”. This comprised (a) the UKFFM Team’s observation note; and (b) notes
of interviews with sources. On 4 August 2016, less than 6 weeks after the
final day of hearing, the Home Office published this report. Responses to a
further direction we made regarding its publication confirmed that its contents
were virtually the same, changes made being only presentational. When we refer
in this decision to the “UKFFM materials”, these are to be taken to include the
report as now published.
101.
One
of the main aims of this mission was described as being to address the issues
identified in the Tribunal’s original directions for the country guidance case
made by the UT in September 2015. It is narrated that the members of the
mission were in Asmara between 7 and 20 February 2016. In addition to visiting
parts of Asmara, the FFM team travelled to Keren, Barentu, Tesseney and Bisha
Mine in the Gash Barka region. The subjects covered by the terms of reference
covered National Service, Demobilisation and Discharge, Evasion/Desertion from
National Service, Leaving the Country, Treatment of Returnees and Position
regarding the UNCOI and other human rights organisations reporting on human
rights in Eritrea
. Interviewees were advised in advance of a number of
“Subjects for discussion” which covered the aforementioned topics but also
including Healthcare/Facilities, including for mental health. The 32 sources
consulted included three anonymous sources, several Eritrean government
ministers and officials including immigration officials, five diplomatic
sources (A-E) plus the UK Ambassador to
Eritrea
, a representative of UNHCR, a
UN Staff member, an international humanitarian organisation, representatives of
the National Union of Eritrean Youth and Students (NUEYS), representatives of
the National Union of Eritrean Women (NUEW), representatives of the National Confederation
of Eritrean Workers (NCEW), two focus groups of Eritrean youths (1 & 2), a
focus group of Eritrean entrepreneurs, a focus group of returnees from Israel,
Sudan, Yemen and Norway, a focus group of artists, a focus group of returnees
(in Tesseney), a focus group of returnees (in Barentev), a focus group of
returnees (in Keren), the Training Manager at Bisha Mine, a representative of
the human resources department at Bisha Mine, Dr Seife Berhe, Director of Andiamo
Exploration Ltd. and an international development organisation.
The new versions of Home Office CIGs, 4 August 2016
Country Information and
Guidance: Eritrea
: National (incl.Military) Service, Version 3.0, August 2016
102.
The latest
version of this CIG, said to be valid from its date of publication on 4 August
2016, together with some new content, contains numerous passages that duplicate
those in the previous September 2015 version. It being the current version, our
summary is more detailed than those of its two predecessors, with particular emphasis
on the entries containing new source information. This CIG sets out a number
of basic facts regarding national service which are stated to include that
national service is compulsory for all persons aged 18 to 50 in Eritrea
with
limited exemptions (2.2.4) and that “sources estimate that between 10-20% of
the population are conscripted, suggesting that the large majority of the
country is not in national service” (2.3.5). At 2.3.6 it is stated that:
“Decision makers must determine whether a person is required to perform national service based on the individual facts of their case. Those who are not likely to be required to undertake national service and therefore are not at real risk on return include those who are exempt:
a. Those who have already completed (and been demobilized from) national service. This may also be evident from their ability to have obtained an exit visa and left the country legally, as conscripts are not granted exit visas.
b. Those who are above national service age.
c. Those who are disabled or medically unfit and therefore have been, or are reasonably likely to be, exempted from national service.”
103. In 2.3.14 note is made that some sources interviewed by the Danish and UK FFMs indicate that the Eritrean government is taking a more pragmatic approach to handling persons who avoid national service, with some individuals who leave illegally and have avoided national service being held only for a short period of time or being simply reassigned to national service duties. It is noted that this may apply to those who work in professions which are in short supply. At 2.3.16 it is observed that a number of sources have also reported that Eritreans who return to the country after 3 years or more abroad are regarded as members of the diaspora, including those who left before or during their national service. After payment of the 2% diaspora tax and signing an ‘apology letter’, they are considered to have fulfilled their national service requirements:
“…The Home Office
FFM of February 2016 met 47 people, most of whom stated that they left Eritrea
illegally before or during national service who paid the 2% Diaspora tax,
signed a letter of apology and returned to
Eritrea
without sanction. Given that
the interviews were arranged by the Eritrean government and the circumstances
of the interviews, this information by itself is not conclusive however it is
consistent with information provided by independent sources that Eritreans who
avoided doing national service and left the country illegally may be able, in
some circumstances, to return without sanction…”
104. Having noted that conditions in national service, primarily military service, are harsh, this CIG has this to say at 2.3.33 regarding pay:
“Unlike those undertaking their national service in the military those in the civil service are not provided with any food or accommodation by the government. As salaries are below the subsistence level, they face severe financial difficulties. In practice, many people take a second or third job or set up a small business to supplement their income whilst doing national service. However the Eritrean government stated in early 2016 that the salaries of conscripts would be raised to a living wage in line with civil service employees. This was effective from July 2015 for certain conscripts graduating from Sawa but is to be rolled out across all conscript groups over 2015 and 2016, and to be paid retroactively. There is some evidence that individuals have been paid the increased wages, although this appears to have been erratically implemented and not yet to be applied across all conscript groups. However low pay, even at rates existing prior to proposed pay increases, is unlikely to be sufficient to constitute persecution or serious harm by itself.”
105. As regards length of national service, this latest CIG notes that sources reported periods of national service from two years to over a decade, but that there is evidence that it is possible for some persons to be demobilised or discharged. The likelihood of release from national service is stated to be influenced by the person’s:
(a) gender - women who are over 27 years old and are, or will be getting, married, or pregnant, or have children are likely to be demobilised or likely able to successfully seek to be demobilised;
(b) occupation – some sources indicate that professions in short supply (such as teachers, geologists and engineers) may find it easier to be demobilised;
(c) area of work – sources indicate that different government ministries, such as the Ministry of Foreign Affairs, may be more willing to demobilise staff than other departments such as the Ministries of Health or Education;
(d) good relations with the person’s commander or reporting officer/manager may make the process easier.
106.
At
2.3.40 it is stated that additionally persons who become medically unfit may
also be temporarily or permanently discharged from national service and at
2.3.41 that a person who is the sole breadwinner of a family may also be able
to be demobilised (it being acknowledged, however, that this is based on a
single example provided by the UK Ambassador to Eritrea
to the UK’s
Fact-Finding Mission Report). At 2.3.42-43 it is stated that:
“2.3.42 There may be additional factors that can increase the likelihood of a person being demobilised, such as the passage of time, those with contacts with government and/or those who are able to pay bribes, and those seeking discharge for economic or family reasons
2.3.43 Decision makers will need to give careful consideration to the length of national service that the person has already served and their prospects of being demobilised or discharged. A long period of national service, even if it is for a decade or more, is not by itself persecution or serious harm “
107. At 2.3.44 the CIG turns its attention to the issue of Article 4 of the ECHR, stating that:
“If it is considered that a person is not at risk of persecution or serious harm but will return to national service, decision makers will then need to consider if there may be a real risk of a flagrant breach of Article 4 of the ECHR which prohibits slavery and servitude and forced or compulsory labour.”
At 2.3.46 it is stated that:
“The onus will be on the person to show that the length and conditions of their national service on return amounts to a flagrant breach of Article 4. Working in the civilian sector in national service, is unlikely generally to amount to a real risk of a flagrant breach of Article 4. Where a person is able to demonstrate that as a result of the open-ended nature of their national service they will face a flagrant denial of their right not to be required to perform ‘forced labour’, they will be entitled to a grant of discretionary leave. Each case will need to be considered on its merits.”
108. Coverage is also given of the diaspora tax as well as the “People’s Army/People’s Militia”.
109. This version contains the following policy summary at para 3:
“3.1.1 National service is compulsory for persons aged 18 to 50 under Eritrean law with limited exemptions. A requirement to undertake national/military service does not, in itself, constitute persecution or serious harm.
3.1.2 The lack of a civilian alternative to national service and the disproportionate penalties for those who refuse to undertake it means that conscientious objectors – in particular Jehovah’s Witnesses and evangelical and Pentecostals Christians – are likely to be at risk of persecution and qualify for asylum.
3.1.3 Evading or deserting from national
service, by itself, is unlikely to be perceived as a political act by the
government. This will, though, depend on the person’s circumstances, including
their actions inside Eritrea
and since leaving the country.
3.1.4 Eritrean law has provisions which punish those who evade or desert from national service with up to 5 years imprisonment. In practice, punishment can be arbitrary and may range from no punishment at all – simply reassignment to another national service post - to several years in prison, where conditions are likely to be harsh and may include ill-treatment. Persons able to demonstrate that they will face a prolonged period of detention are likely to be subject to serious harm.
3.1.5 Persons who have fled from national
service and left Eritrea
illegally may be able to regularise their status with
the Eritrean government by paying the 2% Diaspora tax and signing the letter of
apology. Decision makers will need to consider whether the person has or will
pay the tax, sign the letter and return to
Eritrea
.
3.1.6 Conditions during national service (including the period of military training preceding a national service posting) are generally harsh, although better for the majority who are assigned to posts in the civilian sector, such as in the civil service and as teachers. Some persons, in particular women over 27 who are married and / or have children, may be able to be demobilised / discharged from national service. Persons who are required to do national service in a military posting may be subject to conditions that amount to serious harm. The circumstances of each case will be different and therefore need to be considered on its merits.
3.1.7 The length of national service in law is 18 months but in practice it can be significantly longer and in some cases is open-ended. Open-ended national service, by itself, may not amount to serious harm but where a person can demonstrate that they would face a real risk of a flagrant denial of their right not to be required to perform forced labour under Article 4 of the ECHR, then they may be entitled to a grant of Discretionary Leave. Each case will need to be considered on its merits.
3.1.8 In assessing a case, decision makers should consider if the person will face:
A real risk that they will be punished on return for having evaded / absconded national service.
What is the likely punishment
Will they be required to undertake national service
What conditions they will face during national service if reassigned to a posting
Have they paid or will they pay the Diaspora tax and sign the letter of apology
3.1.9 There may be persons who through a combination of personal factors, including past experience, gender, education and profession, and the person’s link to the government, are not at risk of serious harm. Each case will therefore need to be considered carefully on its facts. Persons able to demonstrate a real risk of serious harm should be granted humanitarian protection unless they are able demonstrate that the risk of harm is for a Refugee Convention reason.
3.1.10 Those who are not at risk of persecution or serious harm may be able to demonstrate that there is a real risk that the length and conditions of their national service will be a flagrant breach of their right to protection form forced labour and, if so, they may be entitled to Discretionary Leave.”
110. In subsequent paragraphs dealing with “Country Information” the CIG identifies significant problems that have arisen over the limits of sources arising from frequent reliance on anonymous sources and observes at 4.1.2 that “information obtained directly from the Eritrean government needs to be treated with caution, and considered against and corroborated with material obtained by other, independent sources.” Reference is made to criticisms of the methodologies of most major sources including the DFFM as well as the two UNCOI reports.
111. The remainder of the CIG contains very detailed coverage of source materials relating to the Eritrean legal framework, size of the military, exemptions and alternatives, conscientious objection, military training, national service postings after Sawa, conditions during national service, duration of national service, discharge/demobilisation and dismissal, law and practice on desertion and evasion and the People’s Army/Militia. As regards duration of national service, it is noted at 12.2.4 that the country analyst section of the Swiss Secretariat for Migration gave the following unsourced summary based on its March 2016 fact-finding mission:
“‘Over the last few years, the Eritrean authorities have announced several reforms to the National Service. Most notably, they promised to limit the length of duty to 18 months starting from the 27th conscription round. This has not been fulfilled yet. National Service remains open-ended and conscription lasts for several years. It appears, though, that a growing number of conscripts who had been deployed in civilian roles are discharged once they have served for between 5 and 10 years. However, no reliable information is available on the demobilization and dismissal of conscripts assigned to the military part of National Service. However, in early 2016, the authorities announced a pay rise in the civilian part of National Service. Apparently, implementation has already started.’”
112.
On the subject
of exemptions and alternatives, this CIG notes at 13.2.4 that sources consulted
during the UK Home Office’s FFM to Eritrea
in February 2016 “also confirmed
that the procedures for demobilisation were opaque and lacked transparency”.
113.
As regards the
treatment of draft evaders and deserters, the compendious treatment given
identifies a number of sources suggesting on the one hand that it will involve
detention and ill treatment, a number suggesting on the other hand that it will
often amount to short detention and/or assignment to (further) national service
duties. At 15.2.18 a reference is made to the very recent report on Eritrea
by the Swiss Secretariat for Migration based on a range of public sources and information
obtained in conversations with interlocutors in Asmara during its own fact
finding mission in March 2016. Extracts from this report include the following:
“‘Deserters apprehended within Eritrea
are usually returned to their military unit or civilian duty and punished. These
punishments are imposed extrajudicially by their superiors. There’s no
possibility of appeal. However, the treatment of deserters appears to have
become less harsh in recent years. Most sources report that first time
offenders are now usually detained for several months. Punishment for deserters
from the military part of National Service is reportedly more severe than
punishment imposed on those deployed in the civilian part. As deserters are not
tracked down systematically, a number of them effectively go unpunished.
‘Draft evaders are usually tracked down in round-ups (“giffas”). Those apprehended are usually detained for some time before starting a military training, which often takes place in camps with hazardous and detention-like conditions. A part of the draft evaders, however, manages to avoid these round-ups in the long run. Sporadically, military units try to individually track down certain draft evaders, particularly those who have been called up already’”.
114. At 15.3.1 the CIG documents sources shedding light on the matter of whether evaders and deserters are perceived as traitors. It states, inter alia, that
“The May 2015 EASO Report, citing various sources, stated:
‘Individuals who leave national service (military and civilian) without permission are regarded as deserters. Most deserters leave either the training centre at Sawa or other military bases without authorisation or fail to return from leave. They then either hide or attempt to leave the country illegally (cf. Chapter 6.4.3). Due to the political and ideological nature of national service, most sources claim that desertion or draft evasion may be regarded by the authorities as an expression of political opposition or treason. Due to the lack of empirical information on the punishment of deserters and draft evaders in the recent years (cf. Chapter 3.8.2), there is no recent information if this is still the case.’”
115. The CIG elsewhere identifies sources that suggest desertion/evasion is not regarded as an expression of political opinion or treason.
Country Information and
Guidance: Eritrea
: Illegal Exit, Version 3.0, August 2016
116. The latest version of this CIG, also said to be in effect from (date of publication on) 4 August likewise contains some new content alongside numerous passages that duplicate those in the previous September 2015 version. It being the current version, our summary of this is also more detailed than that for its two predecessors. In 2.2.3, it is stated that the latest evidence justifies a different approach from the one taken in [133 (iii)] of MO - as the UT’s findings in that case had been made in light of evidence that the Eritrean government in August / September 2008 suspended issuing exit visas, therefore making it very unlikely that any individual could leave the country lawfully. It was now known that this suspension was temporary and the government resumed issuing exits visas. According to Eritrean immigration officials between 60-80,000 exit visas are issues annually - albeit that it may sometimes suspend visa and passport services without warning. Hence:
“While the government restricts to whom it issues exit visas, it remains possible to obtain a visa for certain persons depending on their particular circumstances (see Exit visas). Therefore, as the court found in MA rather than MO, it cannot be assumed that a person left the country illegally if their claim is found to be wholly incredible.”
117. At 2.2.4 this version states that sources published since MO “are broadly consistent with each other and the Tribunal in identifying categories of person who would be able to obtain an exit visa. However, there is evidence that the categories of person who may be able to obtain an exit visa are slightly wider then identified in MO”. The list then set out includes as one of the categories:
“Children aged under 13 (note also that children under the age of 5 are able to exit legally without an exit visa)”.
and adds the observation that:
“ [a]dditionally, there is evidence that women who are over 30 are also able to obtain exit visas to travel abroad (see Exit visas)”
118. At 2.2.6 it is stated that:
“The Tribunal in MO
(and many sources) links illegal exit with evasion/avoidance of national
service and therefore a risk on return. However, there is reason to depart from
the caselaw on this issue as there is now evidence indicating that it is not
illegal exit per se that places a person at risk but the underlying reason why
a person left Eritrea
illegally which may place them at the risk, namely
whether a person has evaded or absconded from national service.”
119. Reference is then made to sources verifying the above statement.
120. At 2.2.8 it is further stated that:
“Additionally, that
illegal exit per se does not lead to risk is demonstrated by the authorities
response to the Diaspora. Thousands of members of the Diaspora return for
varying lengths of stay each year, mostly in the summer months. While many may
have become naturalised in third countries they often use Eritrean documents to
enter Eritrea
(possibly alongside documents from their country of residence), a
number are likely to have left the country illegally and sought asylum /
obtained refugee status. Additionally, over 2,000 Eritreans, some of whom may
have left
Eritrea
illegally and entered Israel illegally, have returned from Israel voluntarily since 2012. Many of the returnees are likely to have regularised their
status with the government of
Eritrea
by agreeing to pay the 2% Diaspora tax
and sign the letter of apology in order to obtain Eritrean documentation and
consular services enabling them to return. However, there is no substantiated
evidence that these persons have been subjected to ill-treatment…”
121. Nevertheless, it is stated in 2.2.9, that as regards why a person left the country without an exit visa,“[i]n most cases it is likely to be because the person has evaded or absconded from national service”.
122. The policy summary at para 3 states:
“3.1.1. Eritreans need official permission
to leave Eritrea
legally. This entails obtaining an exit visa which is stamped
in a passport. Leaving the country without obtaining this is regarded as
illegal exit and, in law, may be punishable with a prison sentence and a fine.
3.1.2 In the country guidance case of MO
the Upper Tribunal held that, apart from some limited exceptions, those who had
left illegally would be at risk on return to Eritrea
. This was because they
would be viewed with hostility by the government and faced arrest, detention
and mistreatment. However, more recent information suggests that the act of
having left the country illegally may not, on its own, result in punishment on
return. It is likely that the reason the person left the country - usually
because they have evaded or absconded from national service – will be why a
person would be of interest to the Eritrean authorities, not the act of leaving
without an exit visa.
3.1.3 A person who
has left Eritrea
illegally may be able to return to
Eritrea
provided they sign
a “letter of apology” and pay any outstanding (2%) Diaspora tax. The Diaspora
tax is a reasonable requirement and a refusal or failure to comply with this
may mean the person is not issued with a travel document to return to
Eritrea
voluntarily, but this would not amount to persecution or serious harm. A person who
has regularised their status with the Eritrean government by
having signed the apology letter and paid the Diaspora tax is unlikely to be at
risk on return.
3.1.4
Decision makers should consider the reasons why the person left illegally and
whether this puts that person at risk. In general punishment solely because a
person left Eritrea
illegally is unlikely however each case will need to be
considered on its individual facts. Where a person is able to demonstrate a
real risk of punishment for having left illegally which amounts to a breach of
Article 3 of the ECHR, humanitarian protection should be granted.”
123. In later paragraphs dealing with the “Country Information” seen to underlie the policy summary, it is noted at 6.1.3 that “the prohibitive cost of passports deters many citizens from foreign travel.” It costs a citizen in national service the equivalent of 40 percent of his or her gross yearly salary to obtain a valid passport” At 7.1.1 is also noted that exit visas cost 200 nakfa and are valid for one month and one trip out of the country. Further paragraphs address the state of the evidence as regards numbers leaving illegally, the shoot to kill policy and numbers of returnees, punishment for leaving illegally and treatment on return and the diaspora tax. In relation to the numbers of returnees it is noted, inter alia, at 10.1.4 that:
“Immigration
officials at a meeting with the UK Home Office’s fact finding mission to
Eritrea
, 7-20 February 2016 (UK FFM), stated: ‘…thousands of Eritreans,
including those who left the country illegally, come back to visit, especially
in summer, to see family, etc. In 2014, 1,538 males and 389 females returned to
Eritrea
. These had left illegally and been away for three years.” In 10.1.8
it is said that “[t]he Population, Immigration and Border Authority (PIBA), of Israel, in correspondence with the Home Office in March 2016 stated that since 2012 2,167 Eritreans
had returned voluntarily to
Eritrea
from Israel.”
124. In the paragraph dealing with punishment for leaving illegally and treatment on return, this CIG sets out sources that indicate that they will routinely face ill treatment but notes at 11.1.6, by reference to the DFFM Report that “there is information to suggest that Eritreans abroad, including those who left the country illegally, are able to obtain Eritrean passports at Eritrean Embassies if they sign an “apology” letter and start to retroactively pay the two percent income tax levied on all Eritrean citizens living abroad.” The extensive sources summarised include more than one that emphasise the difference between the likely treatment of those returning voluntarily and those forcibly returned. Thus at 11.1.26, reference is made to a recent Swiss report arising out of a March 2016 fact finding mission stating, inter alia, that:
“There is hardly
any information available regarding the treatment of forcibly returned per-sons
(sic). In the last few years, only the Sudan (and possibly Egypt) forcibly repatriated Eritreans. As opposed to voluntary returnees, those forcibly returned
are not able to regularise their relation to the Eritrean authorities prior to
returning. The few available reports indicate that the authorities treat them
similarly as persons apprehended within Eritrea
. For deserters and draft
evaders, this means being sent back to National Service after several months of
detention.”
Academics and journalists
125. There are a significant number of articles and academic papers relied on by both parties and we have considered them all, but have selected the following to record because we consider them to be the most relevant to the issues we must consider.
Dr David Bozzini
“National Service
& State Structures in Eritrea
”, 16 February 2012 (Presentation to Federal
Office for Migration, Berne)
126.
The paper is
based on a dissertation prepared by Dr David Bozzini who spent two years in Eritrea
from 2005 to 2007. The results of the dissertation are said to be valid for the
time period of active research namely until 2008 and it indicates that there
have been changes since then.
127.
There is a
certain degree of tolerance towards female objectors and women are able to
travel more freely than men in Eritrea
. They can be subject to roundups.
After the age of twenty-seven, women can regularise their status such that they
are demobilised without ever having joined national service. This route was
introduced in or around 2005. Another way to avoid conscription is through
marriage or pregnancy, but in both cases demobilisation is fragile. Mothers
usually are not remobilised but because of the arbitrariness this cannot be
excluded. There is no systematic practice to remobilise mothers.
128. Most Eritreans have no possibility to obtain exit visas to leave the country legally except demobilised women older than twenty-seven years.
129.
The payment of a
two per cent tax ensures access to all kinds of consular services including the
renewal of identity documents, transfer of money or material to Eritrea
, land purchase in
Eritrea
, heritage matters and legal return to
Eritrea
etc. If somebody
wishes to travel to
Eritrea
who has not paid the two per cent of tax he has to
pay it backdated to the moment he started his exile. People who do not want to
pay the tax prefer not to return to
Eritrea
. There are reports that indicate
that some who return without having paid the tax did not face consequences such
as fines or prison sentences.
Tanja R Müller
“Beyond the Siege –
Tracing Hybridity during a recent visit to Eritrea
” (Review of African
political economy Vol.39, No.133, September 2012, 451-464)
130. Tanja
Müller makes observations based on a two week visit to Eritrea
in October 2011 and interviews she conducted in Tel Aviv with Eritrean refugees.
131. The
article intentionally does not focus on human rights abuses but on the day to
day life of Eritreans in Eritrea
. It is stated that
Eritrea
is often compared
to North Korea but this is far from the truth. Müller mentions those returning
from the diaspora for at least part of the year having been enticed by the
government to buy land. Reference is made to those who leave
Eritrea
illegally and who need consular services and that they are forced to pay a lump sum of
diaspora tax at two per
cent
and to sign a confession.
“But my trip is still
not over, because I don’t get the rights I am entitled to” – what the row over
a Country-of-Origin-Report on Eritrea
reveals about human rights politics (blog
posted on 7 December 2014)
132. Tanja Müller comments on the repercussions following the DFFM Report. She comments that the report is “of shockingly bad quality and little thoroughness and some of its sentences are simply nonsensical or outright laughable”. She also comments that everyone who has any knowledge of PK knows that he has been quoted wrongly and out of context.
133. Reference
is made to the “one dimensional interpretation of Eritrea
” by the human rights lobby
(see para [2] above). The narrative advanced by organisations like AI
or HRW is not recognised by anyone who visits the country. Those who have left
are considered traitors but they do not see themselves as such. There is no
time limit on national service obligations for various population groups.
Media reporting from the
global fringes – Observations from Eritrea
and beyond (blog posted 11 November
2015)
134.
Tanja
Müller considers her recent visit to Eritrea
and the 2015 UNCOI Report which in
her view “took anonymity and confidentiality to a level that makes many of its
statements devoid of context or temporality and thus hard to engage with,
critically or otherwise”. She criticises foreign journalists and academics for
imbalanced reporting. The narrative of
Eritrea
as an unrepentant dictatorship
fulfils an important NGO political function.
Mary Harper
“Africa’s Modernist Enigma”, 22nd June 2016
135.
Mary Harper,
journalist, visited Eritrea
in June 2016. In the article she describes her
visit to Asmara where she spoke with Eritreans who have been in national
service for more than a decade. According to the articles between ten and
twenty per cent of conscripts are in the military and the rest have civilian
roles. One man with whom she spoke had been serving for fifteen years and
supplementing the low pay by selling goods. She met people who have returned
to
Eritrea
from abroad in order to live and work there. One person with whom
she spoke is quoted as saying “
Eritrea
is peaceful, it is safe and there is no
violent Islamic extremism. Of course there are challenges, but this is home”.
She reported that it is very difficult to work out what is going on there in the
light of what human rights groups assert and the United Nations Commission of
Enquiry. She concluded that almost everyone that she met was happy to talk to
her notwithstanding the presence of a camera and microphone. She was not
accompanied by a minder when she openly travelled to
Eritrea
as a journalist
and was not prevented from working there.
Ashish Kumar Sen
“What the UN gets wrong
about rights in Eritrea
”, 7 June 2016
136.
This
report mentions that Bronwyn Bruton, Deputy Director of the Atlantic Council’s
Africa Centre, was interviewed by Ashish Kumar Sen from the New Atlanticist.
Bruton expressed her concern about the UN’s Commission of Inquiry on Eritrea
and their findings that the Eritrean government had committed systemic, widespread
and gross human rights violations. Bruton’s view is that the International Criminal
Court’s (ICC) targeting of African leaders is disproportionate and politically
motivated. There are problems with the Commission of Inquiry’s methodology in
producing the 2015 Report. The Commission refused to consider academic
literature and refused to use press reports. They did not speak with experts
who had recently travelled to the country and refused to speak to UN staff and
western diplomats inside the country. They did not consider the testimony of
many thousands of Eritreans who supported the government and only spoke to
refugees who “self-identified as having suffered violations of their rights”.
She accepts that terrible human rights abuses take place in
Eritrea
but she does not believe that the human rights situation described in the Commission of
Inquiry’s Report is reflective of the reality on the ground. The claim that
Eritrea
maintains a ‘shoot to kill’ policy on the border is “an especially egregious example” and
she said that she had “never heard of any meaningful evidence that would
support that claim, except perhaps in a few, highly militarized spaces along
the border, where
Eritrea
is actively in conflict with its neighbours. But even
there, the evidence seems thin”.
Martin Plaut
“Eritreans Rounded up in Sudan”, 24th May 2016
137.
According to a
report he had received, journalist Martin Plaut reported that nine hundred
Eritreans have been picked up in Khartoum and possibly expelled to Eritrea
. Eight hundred people were deported while getting ready to go to Libya. There are no reports from inside
Eritrea
relating to where the deportees are being held.
Border shootings are increasing on both sides of the border. On May 12th
2016, three Eritreans were found dead near Hamdait (Sudan) from bullet wounds
fired at them by border guards.
“Eritrea
: Naming the Dead and Injured Conscript in Asmara Shooting”, 7 April 2016
138.
From information
obtained from inside Eritrea
, Martin Plaut reports that on 3 April 2016
national service conscripts were shot dead in Asmara as they were attempting to
escape from trucks taking them to the Port of Assab. Twenty-nine conscripts
were killed or injured.
139.
Reference is
also made to the incident (in an article entitled “Shots fired, stoning in Eritrea
’s Capital” of 5th April 2016) on awate.com and on the website
assenna.com.
“Eritrea
Look to Build Mining Sector to Kick-Start Economy”, 26 February 2016
140. This report describes Bisha Mine as being a joint venture between Canada’s Nevsun Resources and the state mining firm EAMCO. Bisha has been “dogged” by allegations from HRW and other groups and former workers about the use of poorly paid workers on national service. HRW and others have described the use of conscripts as “forced labour”.
Edmund Blair
“Eritrea
Won’t Shorten National Service Despite Migration Fears”, 25 February 2016
141.
The thrust of
this article, sent from Asmara, whilst Edmund Blair was there, is that Eritrea
is not prepared to stop forcing its youth into lengthy periods of national service which
drives Eritreans to make the perilous trip to Europe. The Eritrean government
insists conscription is vital for national security in light of the fear of
attack by Ethiopia. Although officially citizens between the ages of eighteen
and forty must complete eighteen months of national service, diplomats and
those who have fled say that this can stretch to a decade or more and that the
government reserves the right to extend time of length of service in periods of
emergency. The article states that
Eritrea
is raising national service salaries
by printing local currency notes to deter people traffickers. In addition it
is investing in mining and other sectors. A western diplomat said that there
was a greater engagement and openness.
Bisha Mines Materials
HRW study entitled “Hear
no Evil: Forced Labour and Corporate Responsibility in Eritrea
’s Mining Sector”, 25 January 2013
142.
Before
us there was also an amount of material relating to Bisha Mines. These included
the 25
January 2013 HRW study entitled “Hear no Evil: Forced Labour and Corporate Responsibility
in Eritrea
’s Mining Sector”. The principal concern expressed in the “Hear no
Evil” Report was the
Eritrea
government’s insistence that the Bisha mine
project, undertaken by the corporation Nevsun, engage Segen Construction
Company as a local contractor. Segen is owned by the ruling PFDJ and “there is
evidence that it regularly exploits conscript workers assigned to it by the
government”. Among the Eritreans interviewed by HRW two said they were
conscripts forced by Segen to carry out construction work during its initial
development; and the report said there was evidence of terrible conditions.
The report expressed concerns about three other overseas mining firms that were
setting up in
Eritrea
and did not appear to heed human rights concerns.
Witness statement from Elizabeth Chyrum
143.
The
materials also included a witness statement from Elizabeth Chyrum of 7 June
2016. It was intended that she would give oral evidence, but at the eleventh
hour she changed her mind about this; her statement, however, is relied upon by
the appellants. Ms Chyrum is a Director of Human Rights Concern – Eritrea
(HRCE) which is a United Kingdom based organisation that works for the
promotion and protection of human rights of all Eritreans through advocacy and
lobbying. She has been looking at the mining companies in
Eritrea
as a result of information received about forced labour. One of the companies which HRCE
investigated in respect of forced labour and conscripted labour is Nevsun
Mining Resources Limited which is said to own 60 per cent
of the Bisha Mine in
Eritrea
. Nevsun Mining Resources Limited is a Canadian
company. HRCE has interviewed former conscripts who have confirmed that they
were subjected to forced labour and harsh working conditions, and that they were
starved and paid very little. HRCE supported three former Eritrean conscripts
who have filed a lawsuit in British Colombia’s Supreme Court accusing Nevsun
Resources Limited of being an accomplice to the use of forced labour, crimes
against humanity and other human rights abuses at the Bisha Mine. The
statement of claim relating to the individuals was produced.
144. As a result of HRCE encouraging former conscripts subjected to forced labour at the Bisha Mine to join the class action, it is said that they have been contacted by a number of people who assert that they worked at the Bisha Mine at various times against their will and under harsh conditions. Ms Chyrum identifies three individuals by initials only who allege ill-treatment and that they had not been demobilised when working for Nevsun at Bisha Mine.
2015 UNCOI Report
145. The 2015 UNCOI Report addresses forced labour at the Bisha Mine which is required to hire Segen and other Eritrean public companies to carry out all of the unskilled labour and basic work construction. Segen was the main Eritrean public company involved in the site work and it sends some skilled workers to Bisha as well as unskilled manual labourers. Segen tried to conceal their status but the majority of the workers were in fact conscripts performing national service and the skilled staff work directly for Segen under the Civil National Service Scheme. The majority of labourers were conscripts whose military units were put at the disposal of Segen by the army. Conscripts, including people who were disabled, were used by Segen to construct the underground network of tunnels for mining operations.
Human Rights Impact Assessment (2015 Audit) – Bisha Mine 5 August 2015
146.
The
respondent produced a Human Rights Impact Assessment of the Bisha Mine in Eritrea
(2015 Audit) commissioned by Nevsun Resources Ltd. The report concluded from
interviews with procurement and human resource managers that progress has been
made with respect to developing a Standard Operating Procedure for including
provisions in all relevant contracts that reinforce the prohibition against
national service workers at Bisha Mine. However, the plan to hire a local
contract manager to coordinate screening and audit activities related to
national service workers has not yet been implemented. The standard screening
procedures requiring documentation of discharge of national service has been
applied to all new contractors or subcontractors before they are allowed on
site. The audit activities at Segen Construction and Transhorn Trucking have
disclosed no evidence of national service workers being used at Bisha mine.
2. Expert Evidence of Professor Kibreab (PK) in summary form
147. It is not necessary for us to set out PK’s evidence in any detail in the main body of our decision as a fuller summary is appended in Appendix III; however, here we summarise the main points.
1).
The
DFFM Report does not accurately represent his views and is flawed generally. It
does not represent the position in Eritrea
.
2).
Draft
evaders or deserters who fled Eritrea
illegally continue to be at risk on
return. The requirement to do people’s militia has effectively increased the
upper limit of draft age which is from 54 to 70 for men and from 47 to 60 for
women.
3).
The
payment of 2 per cent tax and the letter of regret
does not provide protection or immunity; it enables the diaspora to access
consular services. Those returning to Eritrea
are a small proportion of people
who are close to the regime and have been naturalised in another country.
4). National service is indefinite and there is no procedure for discharge/ release or demobilisation.
5). Whilst there is a medical exemption, obtaining exemption on this basis is difficult and rare.
6). The Eritrean government have adopted a stricter approach to the granting of exit visas generally. The categories have narrowed; the lower age for children is now 5.
7). National service is forced labour.
B. ASSESSMENT: THE GENERAL ISSUES
1. Law
The relevance of existing country guidance
148.
The
status of the two existing country guidance cases of MA and MO
has occupied a central place in the arguments before us. This doubtless has
much to do with the fact that these cases identify relatively broad risk
categories whose effect has been that a very significant number of Eritrean
applicants for asylum have been able to show they fall within them. As noted
earlier, it was the publication of the November 2014 DFFM Report that led the
Home Office to announce in a March 2015 CIG on Illegal Exit at paras
1.3.4-1.3.8 that “[t]he most up-to-date information available from inside
Eritrea
– notably the [DFFM] Report” indicated a different view and that
“[c]onsequently, the guidance outlined in MO above should no longer be
followed…” The appellants and UNHCR, by contrast, consider that the guidance
given in MO should be maintained and that, indeed, its risk categories
should be extended.
149. Some of the arguments and counter-arguments ventilated on this issue have a legal hue. Ms Dubinsky on behalf of UNHCR as intervener has submitted that two essential pre-conditions should apply, by analogy to cessation, to the issuing of fresh country guidance withdrawing a previously recognised risk category, or to a finding by the Tribunal in an individual appeal that earlier country guidance recognising a risk category should be disapplied. Those two essential pre-conditions are, she submitted, (1) a requirement of establishing a fundamental change of circumstances; (2) a requirement of establishing a durable and stable change of circumstances. Further, there is a burden on the asylum authority which is seeking to invoke a change since the previous country guidance to demonstrate a durable, stable and fundamental change of circumstances. In developing these submissions she sought to argue that support for her position could be found in what had been said by Lord Brown in Hoxha v Special Adjudicator [2005] UKHL 19 at [63] and by academic authorities including Hathaway and Foster in The Law of Refugee Status and Goodwin-Gill and McAdam in The Refugee in International Law. Mr Rawat, on behalf of the respondent, strongly disagreed with these submissions, arguing that the reference to cessation or a burden of proof was neither helpful nor necessary in the context of country guidance.
150. Ms Dubinsky cited in support of her argument the observations by the former President of the UTIAC, Blake J, in EM and Others (Returnees) Zimbabwe [2011] UKUT 98 (IAC), who at [71] considered that:
“The proposition that a Country Guidance case should provide the “starting point” for a subsequent case that relates to the Country Guidance issue is inherent in the Practice Direction (and its AIT predecessor). Whether the subsequent case is being “set down to review existing Country Guidance” or not, the effect of Practice Direction 12 and section 107(3) of the Nationality, Immigration and Asylum Act 2002 is to require the existing Country Guidance case to be authoritative, to the extent that the requirements in Practice Direction 12.2(a) and (b) are met. This is fully in accord with what the House of Lords (per Lord Brown) held in R (Hoxha) v Special Adjudicator [2005] UKHL 19. If the existing Country Guidance is such as to favour appellants (to a greater or lesser extent), it will in practice be for the respondent to adduce before a subsequent Tribunal “sufficient material to satisfy them” that the position has changed.”
151. Blake J went on at [72] to say that:
”…where a previous assessment has resulted in the conclusion that the population generally or certain sections of it may be at risk, any assessment that the material circumstances have changed would need to demonstrate that such changes are well established evidentially and durable.”
152. Ms Dubinsky reminded us that the latter passage was cited with approval by Maurice Kay LJ (with whom Underhill and Elias LJJ agreed) in MP (Sri Lanka) v SSHD [2014] EWCA Civ 829; Times, July 3, 2014 [21]:
“It goes without saying that extant country guidance which was valid when promulgated should not be changed when the position on the ground remains unchanged. The practice of the UT and, before that, the AIT, was explained by the then President, Blake J, in EM (Returnees) Zimbabwe CG [2012] UKUT 98 (IAC) (at paragraph 72)…”
153. However, as her submission acknowledged, Blake J gave further clarification of the UT’s position in CM (EM country guidance; disclosure) Zimbabwe CG [2013] UKUT 59 (IAC) wherein he stated at [118] that:
“118.What the Tribunal said at [72] of EM is not to be construed as imposing some sort of legal "gloss" on Practice Direction 12, so as to place greater restrictions on a Tribunal making a "later 'CG' determination" than, say, a First-tier Tribunal Judge hearing "any subsequent appeal". It is clear that the Tribunal was not seeking to set a test to be satisfied before Country Guidance could be varied, but merely a means of approaching and evaluating the nature of the changes in the evidence. Where a regime has engaged in persecutory conduct of a particular type even for a limited period, the judge undertaking a subsequent analysis will need to be satisfied that the cessation of the conduct was durable before concluding that either Country Guidance should not be followed or (if engaged in a Country Guidance exercise) that the Guidance itself needed to be amended. There is no rule of law here but simply an application of the precautionary principle relating to the assessment of reasonable likelihood of harm, where the previous assessment of risk was itself based on an unusually virulent and widespread outburst of persecutory activity dating from June 2008, the nature and duration of which needed to be assessed with care.”
154. Moreover, the test articulated by the UT in CM is now well-established in the UT - see e.g. AMM and others (conflict; humanitarian crisis; returnees; FGM) Somalia CG [2011] UKUT 445 (IAC) at [345] - and seems to us to be one designed to reflect in substance the point made by Lord Brown in Hoxha but very properly to decline elevation of it to a rule of law. To seek to elevate the analogy with cessation into a rule of law would in our judgement place the UT in the wholly artificial and untenable position of being prevented from conducting a full ex nunc examination of the latest evidence on the merits. That would be contrary to established authority which provides that asylum appeals in general are decided on up-to-date assessment of risk: see Sandralingam and Ravichandran v Secretary of State for the Home Department [1996] Imm AR 97 (hereafter Ravichandran at p112-113 per Simon Brown LJ) and contrary also to the approach enjoined by Article 4 of the Qualification Directive (Directive 2004/83/EC). Although the UK has not opted into the recast Procedures Directive (Directive 2013/32/EU), it seems to us that Article 46 of the latter represents a clear articulation of the Ravichandran principle and also clear concurrence with the approach taken by the Strasbourg Court when it comes to application of its parallel Article 3 ECHR jurisprudence: see e.g. Saadi v Italy (GC), No. 37201/06, 28 Feb 2008 at [133].
The status of UNHCR Eligibility Guidelines and position papers
155.
Ms
Dubinsky submitted that UNHCR Eligibility Guidelines and position papers on
risk categories in countries such as Eritrea
should be accorded very
considerable weight. In support she cited a number of authorities including
what was said by Sedley LJ in EM (
Eritrea
) [2012] EWCA Civ 1336 at [41],
Lord Kerr’s endorsement of Sedley’s words in EM (
Eritrea
) [2014] UKSC 12
at [71]-[72] and Lord Kerr’s observations in IA (Iran) v Secretary of State
for the Home Department [2014] UKSC 6, 1 WLR 384 at [44] and [49] (with
which Baroness Hale, Lord Wilson, Lord Hughes and Lord Hodge JJSC agreed). The
appellants in their skeleton argument of April 2016 went further and argued
that there should be a presumption that such guidelines should be followed. Particularly
given that the latest UNHCR Eligibility Guidelines in this case were published
(as was MO) in 2011 and that UNHCR, although saying that it continues to
maintain these, has expressed its wish to update them, this would not seem the
strongest case to ask for such guidelines to be given either great or
presumptive weight. In any event, we would simply respond to these submissions
by underlining what was held by the UT in HM (Iraq) [2012] UKUT 409 (IAC) at [277], the latter which were endorsed by the Court of Appeal (per
Elias, LJ) in HF (Iraq) v Secretary of State v Home Department [2013] EWCA Civ 1276 at [44] as follows:
156. We think Mr Rawat was right to remind us that IA was concerned with the approach to be taken by a national decision maker when UNHCR has granted an individual refugee status and the Court made clear in [49] that even in that context “[r]ecognition of refugee status by UNHCR does not create a presumption, does not shift the burden of proof and is not a starting point…”
157. Neither should it be forgotten that these Guidelines themselves do not purport to possess such a special status, stating in the introductory Note, as do all such Eligibility Guidelines, that "it is hoped that the guidance and information contained in the Guidelines will be considered carefully by the authorities and the judiciary in reaching decisions on asylum applications."
158. What we conclude on this issue is that, whilst Ms Dubinsky is entirely right to highlight that UNHCR Eligibility Guidelines and position papers will typically command very considerable respect, they will do so in our judgement because of their typically high intrinsic quality rather than any fixed status.
The status of experts in country guidance cases
159. The appellants’ submissions request that by virtue of the great importance accorded by the Court of Appeal to expert evidence, that “heavy reliance” should be placed on the reports produced for this case by PK. They cite in support S v SSHD [2002] EWCA Civ 539 in which it was said at [29] that “[i]n this field opinion evidence will often or usually be very important, since assessment of the risk of persecutory treatment in the milieu of a perhaps unstable political situation may be a complex and difficult task, in which the fact-finding tribunal is bound to place heavy reliance on the view of experts and specialists”. It was averred that PK is just such an expert and specialist and is “universally recognised as such”. Also prayed in aid were the observations of the ECtHR in NA v UK (2009) 48 EHRR regarding “the authority and reputation of the author”.
160. In view of the fact that the Court of Appeal in the above passage uses deliberately defeasible language (“often or usually”) and that the ECtHR in paragraph [120] of NA likewise viewed the status of evidence about country conditions as a fact-sensitive matter to be assessed by reference to a number of “relevant considerations” (namely “…the authority and reputation of the author, the seriousness of the investigations by means of which they were compiled, the consistency of their conclusions and their corroboration by other sources…”), we do not regard this body of case law as creating any presumption that the reports of recognised experts automatically carry heavy weight. We respectfully endorse the observation made by the UT in MD (Women) Ivory Coast [2010] UKUT 215 (IAC) that “[a] competent expert’s report is always entitled to respect and due consideration but from the point of view of the judicial decision-maker, such reports may sometimes (if not often) amount in the end to just one among other items of evidence which have to be weighed in the balance”. Also pertinent is the guidance given in AAW (expert evidence-weight) Somalia [2015] UKUT 673 (IAC):
“Any opinion offered that is unsupported by a demonstration of the objectivity and comprehensive review of material facts required of an expert witness is likely to be afforded little weight by the Tribunal. In particular, a witness who does not engage with material facts or issues that might detract from the view being expressed risks being regarded as an informed advocate for the case of one of the parties to the proceedings rather than as an independent expert witness”.
2. Methodology and Sources
General Observations
161.
The
respondent has invited us to include in our identified country guidance issues
that of methodology and sources. In doing so she relies largely on the fact
that the case management directions had originally identified the issues to be
covered as including the issue of the DFFM Report. But she also relied on the
wider disputes that have arisen over methodology. In this regard it is fair to
say that all three parties have devoted a considerable amount of time to issues
of methodology, the appellants’ representatives, for example, urging us to find
fault with the great reliance placed by the respondent on anonymous sources
obtained during the DFFM and UKFFM and the respondent urging us to find fault
with the appellants’ great reliance on the UNCOI Reports and certain other NGO
reports based on anonymous sources, especially those said to be asylum seekers.
The Home Office Country Information and Guidance: Eritrea
: National (incl.Military) Service, Version 3.0, August 2016
at paragraph 4 also deals under a sub-heading “Limits of sources”, with some of
the main criticisms made, among others, of the DFFM and UNCOI reports.
162. We have already given our reasons for not treating methodology and sources or assessment of evidence as a country guidance issue in itself and for considering that on several legal issues there is already sufficient guidance in existing case law, but we undertook nevertheless to make specific findings on certain sources. Initially the principal focus was on the DFFM but by the end of the hearing it was clear that there are other sources which have attracted criticism in regard to their methodological basis, not only the DFFM, but also the UKFFM (the two main reports on which the respondent relies, the former now only for its Appendices) and several sources on which the appellants and UNHCR rely, in particular the AI "Just Deserters" Report and the two UNCOI reports of 2015 and 2016. In relation to methodological concerns, it will assist if we deal first with two of the main manifestations of these in the arguments of the parties, anonymity of sources and fact-finding missions. We will then proceed to examine the aforementioned reports in more detail.
Anonymity of sources
163. In their submissions, written and oral, the appellants’ representatives have criticised the reliance both in the DFFM Report and the UKFFM Report upon anonymous sources and urged the UT to take cognisance of the approach set out by the ECtHR in Sufi and Elmi v UK (Application nos. 8319/07 and 11449/07) at paragraph [234]. UNHCR’s skeleton arguments also stated that it “endorsed” the approach of the ECtHR in Sufi and Elmi. We shall deal separately with the status of fact finding mission reports in a moment, but on the issue of anonymity of sources we are disappointed that the appellants and UNHCR should have prayed in aid Sufi and Elmi without due regard to the fact that the Court of Appeal has expressly approved the Upper Tribunal’s stated reasons for differing from this judgment in some respects. In CM (Zimbabwe) v SSHD [2013] EWCA Civ 1303 Laws LJ said:
“I have to say that I deprecate what I see as an attempt to persuade this court to treat the meaning of Sufi & Elmi as if it established something not far removed from a rule of evidence. I would endorse what the Upper Tribunal said at paras 163-165.”
164. We continue to endorse what the Upper Tribunal said in CM (EM Country Guidance: disclosure) Zimbabwe [2013] UKUT 59 (IAC) at paras [157]-[158]:
“Anonymous material is not infrequently relied on by appellants as indicative of deteriorating conditions or general risk. The Tribunal should be free to accept such material but will do its best to evaluate by reference to what if anything is known about the source, the circumstances in which information was given and the overall context of the issues it relates to and the rest of the evidence available.
The problem is not one of admissibility of such material as forming part of the background data from which risk assessments are made, but the weight to be attached to such data. It is common sense and common justice that the less that is known about a source and its means of acquiring information, the more hesitant should a Tribunal judge be to afford anonymous unsupported assessment substantial weight, particularly where it conflicts with assessment from sources known to be reliable. In our judgment it is neither possible nor desirable to be more prescriptive than this, and the task of evaluation of weight is a matter for the judgment of an expert Tribunal that is regularly asked to take into account unsourced data whether submitted by claimants or respondents.”
Fact-finding mission reports: general
165.
Whilst
for the most part the submissions of the appellants and UNHCR recognised that
fact-finding mission evidence has a legitimate role to play if done in
accordance with established guidelines, we cannot ignore that some of the public
criticisms made of the DFFM and UKFFM reports on Eritrea
have verged on generic
a priori arguments to the effect that such missions are inherently
compromised because they are only needed when the country of origin in question
is repressive and it being in the nature of repressive regimes to be closed
societies, they are likely to feature wrongful reliance on anonymous sources,
an unrepresentative range of sources and on government or pro-government sources.
In respect of the UKFFM, the appellants’ submissions complained that the
production by the respondent of the UKFFM materials on
Eritrea
constituted a wrongful “bypass of statutory scrutiny” contrary to s. 142 of the Nationality,
Immigration and Asylum Act 2002. It was submitted that in the context of this
case the fact that the UKFFM materials have not been assessed by the IAGCI
“should give considerable pause for thought before it is accepted as ‘credible
fresh evidence’ warranting adjustment to existing CG, in particular in the
light of the IAGCI’s severe criticism of the DFFM”.
166. We analyse below the main FFM reports under scrutiny in this case, but insofar as the evidence before us contains generic attacks on FFM evidence, our omnibus response is simply to say we see no basis for rejecting FFM evidence as of potential relevance and value in the context of country guidance cases. It is the settled practice of the Tribunal to treat such evidence as of potential value, whether it emanates from a governmental agency or from an international body or an NGO. In EM and others (Returnees) Zimbabwe CG [2011] UKUT 98 (IAC) at [88]-[113] the UT analysed a number of criticisms made of the UKFFM to Harare making clear throughout that they were a source of evidence that had to be assessed alongside the evidence as a whole. A similar approach can be seen to run in other cases such as BK (Failed Asylum Seekers) Democratic Republic of Congo [2007] UKAIT 98 and R (on the application of P (DRC)) v Secretary of State for the Home Department [2013] EWHC 3879 (Admin), 9 December 2013. In general terms, it is always better to be able to assess country conditions if the source material furnishes evidence obtained both inside and outside the country under scrutiny, even when obtaining the former may be fraught with problems.
167. We will deal separately with the argument regarding “bypass of statutory scrutiny” when we address the UKFFM Report.
Fact-Finding missions and the Eritrean context
168. It is more frequent than used to be the case that part of the evidence before the Tribunal in country guidance cases includes FFM reports. Indeed, the evidence of Mr Olsen and Mr Olesen reminds us that such missions only began in the late 1990s. Up to then immigration services in Europe had tended to rely for in-country information on written sources and information provided by embassies abroad:
“The methodology applied and report format has endured since then including putting emphasis on approved notes, i.e. typing up meeting notes and sending them to the interlocutors for correction and final approval.”
169.
It
seems to us that in this respect – transparent presentation of notes of
interviews - the methodology adopted by a number of fact-finding mission
reports in the past decade, including the DFFM and UKFFM reports on Eritrea
, represent a significant advance in the field of COI. Within Europe that advance
has been entrenched by the publication in 2010 of “EU common guidelines on
(Joint) Fact-Finding Missions: a practical tool to assist member states in
organising (joint) Fact-Finding Missions”. We accept there may be reasons why
other reports on
Eritrea
, e.g. those produced by the USSD or AI or by the UNCOI,
do not demonstrate the same completeness and transparency in relation to
sources, but from the point of view of a judicial fact-finding body such as the
UT when essaying country guidance, this feature of recent FFM reports is a boon.
170.
However
the evidence of Mr Olsen and that of Mr Olesen in regard to the DFFM Report
also reminds us that the value of FFM evidence depends on careful prior
preparation aimed to ensure the interlocutors cover a wide spectrum of views
and even then it may be necessary, as was the case with the DFFM Report on
Eritrea
, to rely on a “snowballing” approach whereby one contact recommends
another and so on. The “snowballing” technique carries a real risk that
sources chosen may not be as representative as otherwise. Even the most careful
prior preparation and consultation may not obviate that risk.
171.
We
also consider it important to underline that the controversy that enveloped publication
of the DFFM Report should not be allowed to obscure the value and legitimacy of
efforts on the part of external researchers and analysts to obtain more direct
information from inside a country of origin like Eritrea
. It seems to us, in
the end, that the parties were in agreement on this matter, the appellants and
UNHCR making clear that their issue was with the way in which the DFFM was
approached, not with the mere fact that such a mission was attempted. This
point imports the need to scrutinise with care those criticisms, which, in some
of their public manifestations appeared to adopt the position that because
Eritrea
is a closed society the only body of evidence that could be trusted was evidence
obtained from outside the country. Possibly Messrs Olsen and Olesen lent oxygen
to this misconception by noting that they felt it wrong that the mission did
not include sources from outside the country. As we understand it, the
principal purpose of the mission was to obtain in-country evidence to place alongside
that already available from outside the country. It seems to us that the
presenting difficulty in the Eritrean context was pinpointed, with the benefit
of hindsight, by the Landinfo Report, “
Eritrea
: National Service” which noted
that because the Eritrean government had severely restricted access by
international NGOs to the country, the latter had to base their reports
“largely... on accounts from people who have come to the west and to other
African counties as asylum seekers”. Landinfo, accurately in our view, notes
that this has led to a “paradox that criticism of the accuracy of the sources
has been relatively absent in the various reports published over the years.
Challenges such as reliability, objectivity and accuracy are discussed only briefly.”
It seems to us, therefore, that any criticism that suggests that it is somehow
preferable to confine evidence and sources to those obtained outside a country
like
Eritrea
is quite misplaced. It also seems to us uncontroversial that all
evidence – whether obtained from inside or outside a country - must be subject
to the same rigorous standards. With these initial observations we turn to
examine the respective submissions we had regarding the DFFM Report.
The Danish Fact Finding Mission (DFFM) Report
172. The appellants’ and UNHCR submissions were unequivocal in labelling the DFFM Report “discredited”. The respondent’s position has changed from relying on it in the CIGs of March 2015 to seeking to rely on it in this hearing solely for the evidence disclosed by the Report’s notes of interviews.
173. The appellants do not consider this change in position sufficient and submit that the entirety of the DFFM should be treated as discredited because the evidence of two of its researchers was that the head of mission, Mr Glynstrup, had pressured them and had influenced the contents of the interviews through his manner of questioning. They also relied on the evidence of PK who publicly disassociated himself from the report alleging that the evidence he had given to it had been misrepresented. (The DIS did, of course, publish a revised report in December, which deleted all references to PK’s evidence, but its manner of doing so - retaining the same text, albeit deleted - might not be thought to be accurately described as eliminating all references.)
174. We consider that PK had legitimate cause to complain at the point (on 25 November 2014) when the person in charge of the report went ahead with publication despite having indicated earlier to him that he would be given time to check over whether he was happy with its references to him. However, although PK may not have been given time to check the report (which must have been obvious to those who were responsible for publication), the fact is that he sent an email on 25 November 2014 after it had been published and sent to him as an attachment saying “thank you for this informative and well-written report”. This was in response to having been asked whether he agreed with the report. His oral evidence before us was that he had not opened the attachment, and to this extent he is at fault and to blame for how matters evolved.
175. Subsequently, he complained that the report misrepresented him but he has confirmed to us that what he was referring to was only the purported summary made of his interview recorded in the main body of the report; he was not alleging any distortion of the interview note itself. Whilst in our judgement his sending of an email approving the DFFM Report even though he had not read it was a lapse in professional conduct, we do not count it against him in assessing his expert reports because he has made it abundantly clear that this message of praise was a mistake due to pressure of work and that it does not reflect the view he had of it once he had read it. That still leaves the matter of the meaning conveyed by one particular passage of his interview note. It will assist the reader if we again set out the passage in full:
“In the past two to three years the
government’s attitude towards National Service seems to be more relaxed. It is
now possible for National Service evaders and deserters who have left Eritrea
illegally to return to their country. They must go to an Eritrean embassy and sign
a repentance letter in which they accept any penalty for the offense committed.
In addition, they must pay the two per cent Diaspora tax. Finally, they are
obliged to participate in public festivals in
Eritrea
. In spite of this softer
approach many evaders and deserters still do not dare to return to
Eritrea
, individual circumstances play a role as well. Persons who did not participate in
oppositional political activities abroad and people who are connected by family
bonds or in other ways with government officials or members of the ruling party
would be more inclined to return to
Eritrea
on visits. Gaim Kibreab [PK] was
aware of a few deserters who have visited
Eritrea
and safely left the country
again. These are invariably people who have been naturalised in their countries
of origin”. [We take “countries of origin” to mean “countries of residence”.]
176.
PK
has sought to argue that this part of his text was wrongly interpreted by the
DFFM to mean that there was now a more relaxed attitude towards draft evaders
and deserters, who fled illegally returning to Eritrea
.
177. We find it very hard to read the text the way PK has since said he intended it to be read. The last few sentences do not obviously qualify the first few. At best, the text was ambiguous and, given that he himself had approved it, it was entirely reasonable for the DFFM to infer from it that his position had changed.
178. We dwell on this point here only to explain why we think PK should carry more of the blame for the way in which the controversy developed over the DFFM than he continues to acknowledge. We draw short of suggesting that he must actually have subscribed to the view conveyed by the first few sentences; we are persuaded by the strong reactions he subsequently expressed that he did not mean to subscribe to such a view.
179. As regards the main body of the DFFM Report, we do not need strictly to decide whether it represents a fair summary, since the respondent now places no reliance on it. We would however record that we consider the core criticism made by UNHCR and others - that the main text sometimes takes statements made by interviewees out of context and sometimes ascribes statements to interlocutors that are not contained in the interview notes – well made. By dint of such errors the main text cannot be described as a proper summary. Despite seemingly denying any wish to express policy (“The fact-finding report at hand does not include any policy recommendations”), the main body of the report is as much evaluative as it is descriptive and, insofar as it is evaluative, is significantly flawed.
180. However, with regard to the report’s Appendices setting out the full note of what was said during the interviews conducted with various individuals and organisations identified therein (including PK), we are not persuaded that they cannot be treated as evidence of potential relevance. (The notes are not verbatim transcripts but rather a rendering by the interviewers of what was said). It is true to say that Olsen and Olesen have alleged that in interviews where Mr Glynstrup was present:
“He was rather dominant and he would
always like to try to take control of the interview situation. This was a
major distraction to us. It happened several times that Glynstrup seemed more
interested in having his perceptions of the situation in Eritrea
confirmed by the interlocutors rather than asking open questions and listen to the
interlocutors and reflects on their statements.”
181. It is also the case that although Drs Olsen and Olesen said that they typed up the meeting notes, forwarded those to their interlocutors and received some of the approved notes, “at that very point we were 100% disconnected from the process concerning writing the report” and “we only had the chance to see a few of the approved notes in total before they were included in the report”. However, these two gentlemen did not choose to give evidence to this Tribunal, notwithstanding being requested to do so by the appellants’ representatives. Nor did they offer any explanation for their failure to do so. Mr Knafler has emphasised that we still have their written statement before us which is entitled to significant weight. We do consider some weight should be attached to this statement but cannot accept that this statement – or the various reports of their position in the media and other sources - should be taken to establish that the notes of interviews contained in the DFFM were compromised in any significant way. We of course have limited access to all the relevant documents surrounding their resignations and the criticisms they have made. But the same disability did not confront the Danish Ombudsman. Whilst his report did voice concerns, including about various actions taken by the DIS and other governmental actors in connection with its DFFM Report, it also expressly rejected the complaint of maladministration. In particular (as already noted at [44] above), he found that that “I have no reason to believe that the DIS wished to give the conclusions in the report an untenable expression or put pressure on its staff with this purpose in mind”.
182. In addition, perusing the statements from Drs Olsen and Olesen, neither takes issue with the notes of any of the interviews that were published. We find it very significant that despite saying that at the point when they were shut out from completion of the report they had only received some of the approved notes from interlocutors, neither has suggested that any of the transcripts as subsequently published differ from their own typed notes. We are quite certain that if they had thought there were significant differences they would have said so and they have had many months now in which to say so if that were the case. We also find it significant that despite the blaze of publicity given to the report, not a single interlocutor has come forward and said or suggested the interview notes as published were inaccurate. Certainly some of them were in a position to do so.
183.
Accepting
that the interview notes are likely to be accurate does not necessarily mean
that we accept that the main body of the report accurately reflected their contents
– we have already observed that the main body of the report conflates
description with (flawed) evaluation. We are prepared to accept that Olsen and
Olesen's joint statement raises valid questions about the representivity of the
sources, (both said that Mr Glynstrup prevented them from interviewing
additional sources), even though, as already noted, we do not think it was
wrong in principle for those sources to have been limited to in-country
sources; this was after all a “fact-finding mission” to find out what was
happening inside Eritrea
. We agree that more could have been done to ensure
that the report captured a more complete spectrum of opinion, views, insight
and knowledge in-country. Although these two do not say so in terms, we are
also prepared to accept that they consider Mr Glynstrup’s intervention during
the interviews he attended may have prevented the interviewees from mentioning
all they had to say; that is a feature which certainly calls for a degree of
caution in relying on their contents as a complete account of all they might
have had to say; but such caution does not entirely negate the potential value
of the contents of the interviews as recorded.
184. Similarly we must be cautious, in light of Drs Olsen and Olesen’s evidence regarding the (limited number) of interviews where Mr Glynstrup was present to leave open the possibility that the evidence recorded was not all given in reply to open questions; some of it may have been in reply to closed questions.
185. Drs Olsen and Olesen also criticised the “hazy and unclear” planning of the mission, resulting in normal procedures concerning informing co-partners such as the Danish Refugee Board, but once again that criticism does not go to the potential utility as raw materials of the interview notes.
186. The appellants’ submissions also raise arguments about the potential weight that the DFFM interview records could be accorded given that the interviews with diplomatic sources were inherently limited by (i) the fact that diplomats living in Asmara are prevented by the Eritrean government from having genuine and open access to ordinary Eritreans and from travelling around to see conditions for themselves; (ii) a number of the interviews were with members (or supporters) of the Eritrean government and as such could not be relied upon to present an objective and factual picture of matters such as the conditions of national service and the treatment of returnees; (iii) the report makes no effort to question the particularity and vested interests of Eritrean-based informants; and (iv) some of the quoted information is contradictory or ambiguous or speculative.
187. Taking these objections in reverse order, criticism (iv) and (iii) seems to us to misunderstand the underlying purpose of the interview notes. They are to record what was said, not to put a gloss on their contents or to point out any inconsistencies etc. Such criticisms have some traction in relation to the main body of the report but, as already noted, we place no reliance on that.
188. However, we see considerable force in the criticism that there are more reasons than usual to be cautious about attaching weight to the evidence of the Eritrean Minister of Foreign Affairs, since he had a vested interest in defending the government’s position and reputation, and also the “Regional NGO based in Asmara” who on the basis of the background evidence was also likely to be beholden to the government (this representative’s statement that the country has “no…corruption” is even at odds with the government’s own acknowledgement that corruption is a growing problem). But these are only two of the sources that were consulted.
189.
We
see less force however, in criticism (i) above about the value of the evidence
obtained from western diplomats. So far as concerns the value of evidence from
diplomatic sources at a general level is concerned, we have already noted that
we concur with what has been said on this score in cases such as EM (Zimbabwe). We have commented separately on PK’s opinions about the evidence obtained
from western diplomats, from Asmara, some of which can only be described as
tendentious. But even considering the evidence we have on this issue more
widely, we find nothing in it to indicate that any of the western diplomats
interviewed for the DFFM (or the UKFFM) reports on Eritrea
had an agenda to
distort their evidence so as to portray the situation in
Eritrea
as better than
it was in order to promote a change of view on the part of western government
asylum officials dealing with Eritrean asylum claims. If such criticisms had
been supported by specific instances, for example, of a Western diplomat
voicing ideological views, they may have merited some attention. But so far as
we can tell such criticisms rely purely on a stereotyped portrayal of western
diplomats in
Eritrea
as a class of persons disqualified by their institutional
roles from truthfully describing their own observations and giving their own
opinions based on those observations. PK himself admitted in evidence to us
that his criticisms to this effect went too far and we are confident that this
is a fair description of similar criticisms levelled by others. We note that
neither Mr Knafler nor Ms Dubinsky relied on this particular line of criticism.
190. We attach weight to the view of Drs Olsen and Olesen that “[i]n the case of embassies it became clear that most of their knowledge and anecdotal information stemmed from local staff as well as the other embassies in Asmara...”. At the same time, we think it would be wrong to assume in a generalised way that diplomats and international organisations simply reproduce uncritically what they are told by their informants, and wrong to assume they are unaware, for example, that their own contacts are not representative of ordinary Eritreans or unaware that their staff might include persons who are spies for the Eritrean government. Evidence from these sources is certainly of limited value but is not to be discarded as being inherently naïve or intrinsically ill-informed.
191.
As
regards the reliance in the DFFM Report on anonymous sources, we can see it may
have been possible (as urged by the appellants) for the interviewers to have
provided some further details in some instances, but we consider some of the
demands voiced (e.g. “no indication is given about what information each source
had access to, the degree of authority or level of relevant ‘first-hand’
experience of Eritrea
”) quite unrealistic given the closed nature of the
Eritrean state. We observe that the basic reason for anonymity in the context
of FFMs whose methodology features publication of full notes of interviewees
must be the wishes of the interlocutors and that in any event none of the
sources consulted in
Eritrea
itself is wholly anonymous (each has a descriptor:
“well-known Eritrean intellectual”, “Western embassy…”) so that the reader has
at least some contextualisation.
The UK Fact Finding Mission Report (UKFFM) materials
192.
As
noted earlier, we use the term “UKFFM materials” to encompass not just the
documentation produced to us at the hearing but also the contents of the UKFFM Report
which was published on 4 August 2016. The appellants’ submissions also level a
number of criticisms against the 2016 UKFFM materials, urging that it be found
“no more credible than the DFFM for broadly the same reasons”. They urge that
we should approach them with “huge caution”, there being “significant
methodological concerns about the way in which the sources were identified,
about the impact of the presence of a representative of the Eritrean Ministry
of Foreign Affairs (MoFA) and/or the presence of an affiliated interpreter at
the interviews”. The appellants also raised concerns as to how the interviews
were conducted and how the mission was planned to take account of known
limitations in the gathering of reliable data with Eritrea
.
193. We would first of all note that, no doubt in a conscious attempt to avoid the troubled waters that engulfed the DFFM Report, the UKFFM adheres very closely to the EU common guidelines on (Joint) Fact Finding Missions methodology for fact-finding missions. The terms of reference set out at Annex A identify exhaustively all the topics covered. Annex B identifies precisely what was sent to interviewees in advance. Annex C not only lists the sources consulted but specifies through whom the meeting/interview was arranged; the language of the meeting/interview and the status of notes in terms of whether approved or not. A section headed “FFM Team’s ‘Observations’” gives a purely descriptive account of what they observed. Of the 32 sources listed only four are purely anonymous; all others have some descriptor – e.g. "Diplomatic source…”, “young people”. Some are specifically identified. Furthermore, the notes contain 130 odd pages of verbatim accounts, setting out the questions, the answers and other minor features. We have also had furnished to us a witness statement from Martin Stares which devotes 7 pages to explaining the planning of the mission, its aims, terms of reference, how interlocutors were identified, dates of the mission and itinerary, methodology and how the interviews/meetings were conducted and the process for agreeing the Notes. Most notably, unlike the DFFM, the UKFFM attempts no executive summary – which has indisputably been the main target of the criticisms made of the former. This is an unpromising start for a submission (by the appellants) that the UKFFM evidence is “no more credible than the DFFM for broadly the same reasons”.
194.
We
have already noted when analysing the DFFM Report that even discounting
unwarranted reliance on stereotypes of “Western diplomats”, there are certain
limitations to the potential weight we can give to evidence from the diplomatic
community generally in Eritrea
. Whilst we think these also apply to the
diplomatic sources identified in the UKFFM materials, we find that the evidence
of the HM Ambassador merits somewhat more weight because we know more about it
and it does indicate that he has been able to move around
Eritrea
to a
significant extent, visiting Massawa, Tesseney, Barentu, Bista, Keren and Adi
Quala and in the course of these visits had spoken to “ordinary people,
business people, ministers and officials. I regularly go hiking at the
weekends in villages around Asmara and can and do speak freely to Eritreans. I
meet there, and also in social venues such as coffee shops or the markets. It
is not unusual for people to start conversations with me in the street”. If he
had any reason to think that such conversations were monitored or compromised
by the Eritrean government, we are confident he would have said so.
195. This brings us to the point noted earlier when outlining the legal points relied on by the appellants’ representatives, namely the submission that the UKFFM materials cannot be relied on because there has been a “statutory bypass” of the process by which such reports are monitored by the IAGCI. (The Independent Advisory Group on Country Information (IAGCI) was set up in March 2009 by the Independent Chief Inspector of Borders and Immigration to make recommendations to him about the content of the Home Office‘s COI material.) We are not persuaded that the UKFFM evidence is diminished by the fact that it has “not resulted in any report by the SSHD, properly evaluating what if any evidence should be placed on the UKFFM material.” Even if there had been such a report before us, our primary interest in the UKFFM materials would have been (no less than it is now), as a record of interviews conducted and of what was said, not on any IAGCI commentary on what else might have been said or inquired about. We may have learnt something more from such a commentary, but the notion proposed to us - that a judicial fact-finding body should ignore or treat as tarnished in value existent UKFFM evidence simply because it has not yet been through the filter of an advisory body to the government on COI – is one we find frankly absurd.
196. We also consider that certain other appellants’ criticisms of the UKFFM materials misunderstand their status. Of course, the mere existence of verbatim interviews, approved by the interviewees, does not render them “ipso facto credible and relevant”; they “require proper evaluation”. But, that is far from being a reason for not treating them as “raw material” evidence in the first place.
197. The appellants contend that the value of the UKFFM materials is reduced by the reliance on 27 “anonymous sources” and absence of information that enables the Tribunal to assess the reliability of a source such as information about the nature of a source’s operation in the relevant area. Criticism is also made of the vague descriptions of anonymous sources; the lack of a persuasive justification for anonymity; and the exclusive/preponderant use of anonymous sources. As regards the significant reliance on anonymous sources, we have already explained why we consider that this feature reduces but does not extinguish the value of such evidence. Furthermore, only three sources were wholly anonymous. Whilst we consider that in relation to these (and also some of those where some descriptor was given), more information could have been provided, this lack does not negate its value entirely.
198.
Insofar
as the appellants seek to argue that they have been handicapped from being able
to comment critically on the UKFFM materials by a failure on the part of the
Secretary of State to disclose “highly material underlying documents including
correspondence with the Eritrean Government and the original interview notes”,
the Tribunal has already explained why this argument lacks substance in its
Interlocutory Judgment: MST and others (Disclosure – restrictions – implied
undertaking) Eritrea
[2016] UKUT 337 (IAC) at [10] . We would only
reiterate here that we have not been presented with any evidence to suggest
that the published interview notes are an inaccurate record. We are satisfied
that the explanation set out by Mr Stares of the underlying methodology plus
the “Observations” document, plus the further information provided by Mr Rawat
suffices to enable us to evaluate that record.
199. Indeed as a result of the information about methodology provided by the SSHD, the appellants and UNHCR have been able to identify and highlight that of the 32 sources, 17 were arranged by the Eritrean Ministry of Foreign Affairs (MoFA) including all eight focus groups regarding whom (as a result) “it can be assumed, [were] identified as suitable to participate in discussions with the FFM, by the MoFA”. We concur with the appellants that those involved cannot be assumed to be independent witnesses and may have been simply acting as the “mouth-piece” of the government. We know from a number of country reports that the Eritrean government is anxious to regulate and control access by foreigners and is extremely sensitive to potential criticism. Press statements put out by Eritrean government representatives disclose that they consider that there has been a concerted campaign mounted against them by various UN organisations, NGOs and others to portray it as despotic. The government’s decision in 2015 to publish a scathing attack on UNHCR’s 2011 Eligibility Guidelines – summarised at [32]-[33] above, is just one such example. As a result, the notes of interviews conducted with Eritrean government representatives or ruling party members or supporters or persons who may be beholden to the Eritrean government must be treated with very considerable caution. They are helpful to us in understanding the approach of government representatives and supporters, but we do not consider, without more, their contents should be relied on in any significant way.
200. The evidence we have indicates that several of the civil society organisations, the NUEYS, NUEW and NCEW, are affiliated with the ruling PFDJ and in any event cannot be considered to be in a position to speak freely. We concur too with the appellants’ and UNHCR’s point regarding the fact that a representative of the MoFA sat in on 12 out of the 32 interviews and was present during the interviews with the Minister of Justice, Minister of Health, Minister of Finance, Immigration Officers, the Head of Political Affairs, the Regional Governor of Gash Barka and the representatives of the NUEYS, NUEW and NCEW, 7 of the focus groups; and that an interpreter from the MoFA was also present during the interview with the focus group of returnees from Tesseney and took an active part in discussions. From Mr Stares’ statement we learn that notes of the meetings/interviews with the 8 focus groups were sent to the MoFA. Whilst we do not know whether the participants were told in advance that this would happen, in our judgement this tends to confirm that those participants knew that what they said would become known to the MoFA. We agree that in such circumstances it cannot be assumed the participants were able to speak freely.
201. We deem less significant that of the 31 interviews, only 20 had been approved by the interviewees, with the remaining 12 having been sent but not yet approved. This factor does reduce the value of the interview notes somewhat, but does not extinguish their potential value as evidence.
The two Amnesty International Reports (AI Report on AA and “Just Deserters”, the two UNCOI Reports of 2015 and 2016 and the witness statement from Elizabeth Chyrum
202. A recurrent theme of the respondent’s submissions in this case has been that there are as many if not more methodological problems with key parts of the appellants’ case than those alleged to infect the reports relied on by the respondent. In this regard she has taken particular issue, inter alia, with the evidence of PK, two AI Reports and the two UNCOI Reports and a witness statement from Elizabeth Chyrum. We deal separately with PK’s evidence below at [228]-[240] and have also summarised Ms Chyrum’s – see [143]-[144] and [224] - but must now turn to assessment of those other items.
203. As noted earlier, the AI Report on AA and “the Just Deserters” Report rely on the same core data, namely reports from 72 interviewees who were interviewed between July 2014 and July 2015, all “recently arrived asylum seekers” (para 84 of the AI Report on AA). Because the “Just Deserters” Report has been published we will focus primarily on it, but note that we have taken account of the contents of both in full.
204. The respondent’s criticisms of the methodology of “Just Deserters” are essentially sixfold: (i) it relies heavily on the accounts of asylum seekers, i.e. persons whose accounts have not (so far as is known) been tested by a decision-maker or a tribunal; (ii) it does not set out who “the range of sources and interlocutors” used to identify the individuals to be interviewed; (iii) it is not made clear who were the ‘Eritrean activists’ from whom information was also taken, or what independent views they offered or what questions they were asked on what topics; (iv) a large number of the propositions in the reports are completely unattributed, and do not specify the actual number or percentages of the interviewees supporting them; (v) the methodology used in the interviews is not made clear – whether they were all asked the same questions, whether they were interviewed alone, whether they were asked open or closed questions, etc; and (vi) in at least two instances the text from two sources is identical or near-identical (one instance being “Filmon” on page 26 and “Yonas” at page 45).
205.
We
find that to a varying extent the respondent is right to draw attention to the
above features of the AI reports. In relation to (i), we can understand that
AI may have wanted to focus on persons who had recently left Eritrea
so as to give an up-to-date picture. We also accept Mr Knafler’s point that it would be
wrong to apply stereotypes to asylum seekers; it would be utterly wrong, for
example, to assume they have a vested interest in lying. At the same time, AI
is fully aware of the concerns expressed by the Upper Tribunal in a number of
cases about sources that rely heavily on asylum seekers’ evidence which has not
been tested and, as PK’s own writings attest, analysts cannot expect
decision-makers to assume such evidence would stand up to judicial scrutiny.
206. We accept that AI considers that it seeks to verify the evidence of witnesses “to the greatest possible extent” (see above at [65]) but in the absence of any indication whatsoever of such evaluation having been applied, it is impossible to gauge what that means in practice; and it is clearly difficult for any organisation seeking to give absolute priority as AI says it does to protecting the anonymity of witnesses and reassuring them they can give evidence safely, to pursue lines of questioning that might be perceived as expressing doubt.
207. Considering matters in the round, we fail to understand why no interviews at all were conducted with Eritreans whose asylum claims had been found to be truthful by national decision-makers and/or whose claims have resulted in refugee status. If none were available that should itself have been explained.
208. As regards (ii), once the decision was made to rely on interviews of asylum-seekers only, we do not think it matters very much that the reports do not say more about the range of sources who identified them to AI; it is highly unlikely that there would be 72 interviewees who all know each other or reflected just one or type of case. As regards (iii)-(v), we consider that the missing information they identify would have helped enhance their possible value as evidence, although this does not negate it. We note that it is not AI’s practice to disclose or publish even anonymised interview notes. That is clearly a matter for AI, but in a world in which the corpus of available evidence may include FFM reports that do include such notes, they cannot complain if this comparative lack of transparency is seen as a shortcoming. We note that if we had seen such notes we may have been in a better position to make sense of the worrying allegations made by the respondent in (vi) above in relation to the apparent reliance on identical evidence from two sources. It is possible in the light of such fuller evidence we may have been able to establish whether these were isolated examples. As it is, we cannot rule out that the reports relied on may contain other apparent errors of this kind. Despite Mr Knafler telling us on the second day of the hearing that those instructing him hoped to receive an explanatory note from AI regarding the two examples identified by Mr Rawat, no such note materialised, nor any explanation for its non-production. AI, we remind ourselves, produced its Report on AA for the purposes of this hearing.
The UNCOI Reports 2015 and 2016
209. The respondent has also assailed the methodology of the two recent UNCOI Reports, noting that critics of the first include: Lifos, who have queried whether their temporal scope is overbroad (trying to assess the performance of the Eritrean state from 1991-2015) and stated that “from a source-critical perspective [it] has some weaknesses”; Bromwyn Bruton, Deputy Director of the Atlantic Council’s Africa Centre; and Dr Tanja Müller who has written that the report “took anonymity and confidentiality to a level that makes many of its statements devoid of context or temporality and thus hard to engage with critically or otherwise”.
210. The respondent’s main criticisms in full are that: (i) the overbroad temporal scope renders the report’s methodology “entirely opaque”; (ii) the Commission does not set out how the interviewees were selected or what steps were taken to protect against interview bias; (iii) the Commission does not explain what percentage of respondents were asylum seekers and does not grapple with the issue of whether the evidence of asylum-seekers can always be taken at face-value; (iv) the Commission does not give any detail about the nature and methodology of the questioning, who carried out the interviews, whether the questions were open or closed; or whether interlocutors were alone or not; (v) it is not clear how written submissions were checked or verified – only selected extracts of interviews were made available and there is use of paraphrase; (vi) the vast majority of propositions are supported by very few sources.
211. Despite the respondent having set out the criticisms of the Commission in her written and oral submissions, the appellants’ written submissions contain no rejoinder to them. In oral submissions Mr Knafler made no reference to them in respect of the 2015 UNCOI Report. He submitted that in the second report fresh evidence was relied on.
212. UNHCR’s supplementary written submissions do not address the respondent’s criticism of the 2015 UNCOI Report, but (as do her supplementary submissions addressing the August 2016 versions of the CIGs) they do address the status of the 2016 UNCOI Report in the context of criticism made of the 2015 Report. These submissions observe that whatever force the criticism of the 2015 UNCOI Report for its wide temporal scope might be thought to have, that critique is not applicable to the 2016 UNCOI Report, since at [74] the latter report noted that:
“All the witnesses and other evidence cited in subsections 1-10 of this section of the report on current human rights concerns detailed violations that took place between 1 June 2014 and the date of issuance.”
213.
The
2016 UNCOI Report, UNHCR points out, also identified that its pool of
interviewees was drawn from thirteen countries and the Commission
also spoke with experts, diplomatic staff of third countries currently working
in Eritrea
, foreign journalists who recently visited
Eritrea
and other UN
agencies and NGOs. UNHCR submitted that both the 2015 and 2016 reports identified
that their interlocutors included not just victims but former members of the
Eritrean government and commended the care with which the 2016 UNCOI Report had
reviewed the 44,267 responses to its call for written submissions (which came
from 39 countries), noting that it considered a randomly selected sample of
2,250 of these respondents and contacted the author of each one to verify its
authenticity. The UNCOI had correctly noted the highly generalised nature of
the assertions and denials. UNHCR considered the Commission had fairly assessed
the limited value of the majority of these responses that were critical of the
2015 UNCOI Report. According to UNHCR “the thoroughness of the 2016 Report’s
analysis affords a striking contrast to the Danish and UKFFM Reports.”
214.
We
are wary of reaching a definitive view regarding such criticisms. The fact
that the 2016 UNCOI Report records that most of the 44,267 responses it had to
the 2015 Report, coupled with the information contained in a Shabait press
release dated 23 June 2016 (‘Eritrea
–Ministry of Information’) that when it was
officially launched in Geneva, a protest demonstration against it was said to
have been attended by 6,000 persons, is a vivid illustration of the strong
feelings engendered by all reports on
Eritrea
that have implications for the
treatment and processing of Eritrean asylum seekers and the international profile
of the Eritrean state more generally. We are also conscious that our focus is
very different from that of the UNCOI Inquiry. We are not tasked with deciding
on the nature and extent of human rights violations in that country over a 25
year period and it would be arrogant in the extreme for a domestic tribunal
dealing with a country guidance case focusing on risk on forcible return to try
and pass judgement on a large-scale international inquiry which has taken
several years and involved a prodigious amount of work. On the other hand, we
cannot avoid identifying certain difficulties posed by these two reports that
impinge on our own task and we agree with the respondent that we cannot apply
different standards from that we apply to e.g. government Fact-Finding
Missions, just because the report is carried out by UN officials.
215.
We
venture no criticism of the Commission for its wide temporal scope, since that
was clearly the remit it was given, but this feature does make it very hard to
ascertain the precise evidential basis of the 2015 Report for its assessment of
the situation in Eritrea
in 2015. It is a pity that the 2015 Report tells us
so little about how the interviewees were selected. We note that unlike AI's
“Just Deserters Report”, the UNCOI interviewees are said to include refugees as
well as asylum seekers, which potentially reduces the scope for concern about
reliance on untested evidence. Yet the report’s failure to identify how many
non-asylum seeker “victims” were interviewed does not assist.
216. We are very conscious that the UNCOI makes very clear that it is not a judicial body, but at the same time it does state that it has applied rigorous standards and it does purport to apply international law principles and for this reason we would have hoped that the report’s methodology would have given more context regarding such matters as whether questions were open and closed, whether anyone else was present etc.
217. The fact that the 2015 UNCOI Report only includes extracts from interviews is a feature that causes difficulties in being sure they link to different witnesses. The respondent also makes fair points in analysing the extent to which key propositions in the 2015 Report are only supported by limited sources.
218. As regards the 2016 UNCOI Report, it is indeed much more helpful for our purposes in identifying that although still forming part of an inquiry into the past 25 years, one of its specific purposes is considering whether there had been any significant changes since the first report: see [54] above. On many key issues relating to military/national service the level of detail and cross-linking to primary and secondary sources is extraordinarily impressive and conveys to us that the authors have refused to rely on generalisations and have eschewed the temptation to simply regurgitate materials from elsewhere. We take note that it relied on more than 830 sources of information.
219.
We
do have concerns nevertheless about the way that the 2016 UNCOI Report
responded to the fact that the “bulk” of their 44,267 submissions respondents
expressed views critical of the 2015 UNCOI Report. We found very helpful the
response of the Special Rapporteur to our further directions sent in late July
to questions regarding this concern. Although making clear that since the
Commission has now completed its task she is not in a position to make detailed
statements concerning its methodology, her responses shed further light. Nevertheless
they do not entirely allay our concern. The authors may well be right in stating
in the report that a good number of these respondents had not read the 2015
Report and were orchestrated by pro-Eritrean government actors (that was a
point reiterated by the Special Rapporteur in the 15 August letter), but that does
not wholly explain why, as a result, none of these individual responses are
referred to in the report itself. The Report details that it considered a
randomly selected sample of 2,250 of these respondents each of whom was
“interviewed to verify the authenticity of the submission”. In the 15 August
letter from the Special Rapporteur it is explained that the Commission did not have
the resources to review each and every submission and that it took steps to
ensure that the sample group of 2,250 was selected so as to cover all
languages, geographic areas and gender and that it then selected 500 writers
located in 126 countries to contact individually and that - “although invited
to provide further information - “[n]ext to none added the type of factual
detail that would have permitted consideration in the findings on international
crimes and human rights violations” and “[n]one of those contacted chose to
discuss their own personal experiences in the national service, although most
stated generally that national service in Eritrea
is a necessary response to
the numerous threats that
Eritrea
faces”. This confirms what was stated in the
report itself about the highly generalised nature of the assertions and denials
and the fact that “next to none of the authors referred to their own military
national service, the conditions of their military national service or the
length of their service” and that many were vague about their own reasons for
leaving
Eritrea
. The letter further observes that “[none] of them described
witnessing a situation in which human rights violations had been said, in the
first report to be occurring…”. The Special Rapporteur
further stated: “Had any of the writers provided, whether in writing or in the
sample phone-calls, any substantive information with respect to the
crimes/human rights violations at issue (including to state that they had
concrete evidence that these were not occurring), the Commission would have
made follow up contact with the writers and asked them whether they would be willing
to speak formally as witnesses. Their evidence would then have been assessed in
sections III-IV of the report. There were in fact, at the request of a handful
of those who the Commission contacted by phone, some further follow up calls
but these still did not elicit relevant information of substance”. However she also
accepts that the Commission “chose not to ask specific questions, including
concerning military service” and that the interviews were “specifically directed
at understanding what weight could be attributed to the written submissions,
given the appearance of a coordinated campaign”. Given that presumably some at
least of these respondents were people who have lived in
Eritrea
since 1991, they
must all or many have performed some period of national service and if asked
about this may have been able to provide concrete information going to the
issues within the Commission’s remit, potentially affecting for example the
extent to which human rights abuses were systematic. We entirely understand the
dilemma facing the Commission in terms of its limited resources, but having
sought further submissions and then received some 44,267 submissions, even
eliminating those found to be formulaic or coerced, we do not think they could
so easily be discarded as potential sources of evidence relevant to the issue
of the military/national service system in
Eritrea
. The responses received are
not identified anywhere in the report except by way of a summary and an analysis
of “common themes” at [48]-[55].
220. The treatment of these submissions is in sharp contrast with the specific use made of “witness” evidence that reflected adversely on the Eritrean government, which is used throughout to corroborate various findings made in the report. It seems to us that the same methodology should have been applied to all the sources, whether they were “respondents” or “witnesses”. To underline the point we have made already, the letter from the Special Rapporteur accepts that the efforts made to follow up by contacting a sample of respondents did not include asking them if they were able to describe their own military/national service experiences. If they had it may have yielded potential evidence from at least some. This would not have prevented the Commission from identifying that it considered any such further evidence received in response to be partisan or otherwise deficient.
221.
It
is surprising that despite noting at [22] that most respondents stated that
they visited Eritrea
only occasionally and that many stressed the general sense
of calm and order in Asmara, the Commission’s only expressed response to this
information was to observe that the types of human rights abuses committed in
Eritrea
are not committed on the streets of Asmara. We doubt that all of these writers would have
regarded calm and order on the streets as proof that all was well throughout
Eritrea
.
222.
We
would add that we find it at least curious that pursuant to Council Resolution
26/24 the Commission of Inquiry should include as one of its members the
Special Rapporteur, someone who in proper exercise of her remit for that post
(she was appointed in October 2012) had already gone on record on numerous
occasions as someone highly critical of Eritrea
’s human rights performance.
For her to give her own evaluation was precisely what the UN would expect of
such a Rapporteur. However, in September 2014 the Human Rights Council
appointed her to the Commission of Inquiry. We are sure there were worthy
motives behind this action, including the value of the Commission benefiting
from her existing expertise. We also take judicial notice of the UN’s Rules of
Fact-Finding Procedure for UN Bodies Dealing with Violations of Human Rights
adopted in 1973 and the Belgrade Minimum Rules of Procedure for International
Human Rights Fact-finding Visits, approved by the 59th Conference of
the International Law Association, held in Belgrade in 1980 and the Guidelines
on International Human Rights Fact-Finding Visits and Reports (The Lund London
Guidelines) 2009 and the fact that more than one previous UN commission of
inquiry has included a Special Rapporteur member. But from a procedural
perspective it does open the report to criticisms as to its impartiality (the
Lund London Guidelines, for example, at para 8 state that “The mission’s
delegates should comprise individuals who are and are seen to be
unbiased”(emphasis added)). In the judicial context it would ordinarily be
expected that anyone appointed to an inquiry had not previously reached any
publicly expressed view on the issue in hand, so that the public can be assured
they approach their task with an open mind.
223. At the same time, we consider it extremely important not to allow the difficulties we have just identified to blur perspective and we do not consider that they significantly undermine the fact that the Commission’s findings were based on a very substantial number of first-hand accounts. It cannot be gainsaid that the two reports taken together, represent a large-scale, sustained and intensive effort to detail and evaluate all relevant aspects of the Eritrean state on the basis of substantial, first-hand evidence.
Witness statement of Elizabeth Chyrum
224. We have summarised the evidence of Ms Chyrum contained in her witness statement regarding Bisha Mines and the evidence generally relating to Bisha Mines. Ms Chyrum did not give evidence, although initially it was her intention to do so. We accept that there is evidence of exploitation generally in the mining industry and that a number of conscripts have complained to Ms Chyrum of being subjected to forced labour, three of whom have joined a class action. Others are being encouraged to do so. We are in no position to assess their claims and as Ms Chyrum would not give evidence we can only treat it in the same way as much of the other background evidence. However, the evidence overall does establish that some conscripts may be subject to forced labour in the mining industry and we will return to this when we assess forced labour.
The Home Office Country Information and Guidance (CIG) publications on 4 August 2016
225.
As noted earlier
on the same day that the Home Office published the UKFFM Report on Eritrea
, it
produced new versions of its two CIG notes on
Eritrea
: Country Information and Guidance:
Eritrea
: National (incl.Military) Service, Version 3.0, August 2016 and Country Information and
Guidance:
Eritrea
: Illegal Exit, Version 3.0, August 2016. Having learnt
of their publication whilst still deliberating on this case the Tribunal
decided to make further directions affording the parties the opportunity to
make submissions on their significance and relevance. In response the
respondent pointed out that in large part save for very limited exceptions
relating to a journalistic piece from Mary Harper, the Landinfo mission of
February 2016 and a Swiss fact-finding mission of April 2016 report (the latter
which had only been available in German at the hearing stage) these relied on
the same body of sources already before the Tribunal. The response by the
appellants questioned the timing of their publication, which “appear to be a
direct response to anticipated guidance on key issues ventilated during these
proceedings”. They drew the Tribunal’s attention to the sections of the CIG
which have undergone considerable revision or editing when compared to the
previous version relied upon during the proceedings submitting that “[n]o
explanation (or evidential basis) is proffered by the SSHD as to why critical
passages from v2 have been omitted from v3.” UNHCR’s response expressed concerns
about the timing of the new CIG versions “less than a month before the anticipated
date of the judgment.” These concerns were said to be “reinforced by the
problems in the new CIGs of unexplained reliance on sources suggesting positive
changes in preference to more critical sources; selective citation of sources
in the CIGs; and the heavy reliance on the methodologically flawed Danish and
UK FFMs.” Among the points raised was that the new CIGs did not properly
reflect the balance of the evidence relating to the likely perception and
treatment on return of draft evaders/deserters and the serious consequences for
persons failing to comply with their obligation to serve in the people’s
militia. UNHCR also submitted that the criticisms made in the new CIGs of the methodology
of the two UNCOI Reports were unwarranted.
226. We have taken fully into account the two new versions of the CIGs and the parties’ further submissions on them. We make no criticism of the respondent for acting to publish the new versions of the CIGs, to align with the publication of the UKFFM materials (the latter which she had made clear at the hearing had been served on the Tribunal as soon as they became available, even though publication would take a little more time). That said, it is unfortunate that the respondent did not make clearer on the last day of hearing (on 20 June) that the Home Office planned to publish them within weeks. Given the sequence of events we wholly fail to understand the basis on which the appellants sought to submit that their publication was designed to “head off” the anticipated guidance from the Tribunal. The respondent could not have anticipated that we would decide to invite submissions on them.
227. Beyond the above observations, we do not propose to set out our assessment of the responses we received. It suffices to say that we have taken them into account. We have already summarised the contents of the new CIGs and, in line with the general structure of our decision we refer to them and/or submissions made regarding them as and when appropriate.
C. ASSESSMENT OF THE EVIDENCE OF PROFESSOR KIBREAB (PK)
228.
PK
gave evidence to the Tribunal in MA and MO and in the latter case
the Tribunal stated that “[l]ike the Tribunal in MA, we consider that PK
should be considered as a serious expert on country conditions in Eritrea
”;
and, whilst not accepting every aspect of his evidence, it concluded that his
evidence generally should be accorded serious weight ([93]). Unlike the
position in MA and MO the respondent in the present case has
mounted a great many criticisms of PK’s evidence and it is fair to say that as
a result of the events surrounding the publication of the DFFM Report and the
use made subsequently of it by the UK Home Office, PK has taken a very public
position regarding this report and regarding the policies of the governments of
Denmark, the UK and other Western states concerning the treatment of claims by
Eritrean asylum-seekers.
229. One of the principal criticisms levelled by the respondent is that PK let himself become too personally involved in the DFFM controversy to be able to give independent evidence regarding it and that this has carried over into his approach to the UKFFM. We regret to say that we see a certain force in that criticism.
230. As noted when assessing the DFFM Report, we find that PK must bear some of the blame for the way that the public controversy unfolded after the DFFM publication because he had not checked through a draft of it as he was asked to do and had sent an email nevertheless to the DIS describing the published report as “informative and well written” when he had not even read it. He had also approved the note of the interviews he had given to the DFFM researchers despite the fact that it included a passage which was clearly capable of being read as conveying that he believed there had been a relaxation in the policy of the Eritrean government to draft evaders and deserters who had left illegally: see above at [175]. We accept from his evidence to us that at the relevant time he was under very considerable pressure in his university work, but instead of immediately recognising and acknowledging that this passage was not what he meant to convey, he blamed the DIS for distorting what he had told them. There were indeed serious errors in the DFFM, which we have analysed above. We accept PK’s evidence that its head did not give him time to check the report having initially said he would do so, but we find that PK did not help the terms of the subsequent public debate by failing to make clear at the time his own responsibility for not checking his note of interview and the rest of the report. His use in his early public statements of the language of “betrayal” did not adequately explain his own errors and did not assist the level of public understanding and did not exemplify the behaviour we would expect of an experienced academic expert.
231.
It
seems to us that the extent to which PK had let himself become personally
embroiled in the public debate over the DFFM and the UK Home Office response to
it is borne out by the language that he employed in the critique he wrote of
the DFFM immediately after it was released and in the AA Report of September
2015. There are passages in this September 2015 Report which we can only say
are untypical of reports we have read from him over the years. More than one
passage says in effect that the diplomatic community in Asmara has a vested
interest in painting an untrue picture of conditions in order to stem the flow
of Eritreans seeking asylum (see e.g. A1/8). His commentary of March 25, 2015
entitled “Some Reflections on the UK Home Office’s Country Information Guidance
Eritrea
: National (incl) Military) Service and Illegal Exit, March 2015” and
his revised report of 4 April 2016 for this case contain similar assertions;
and the latter endorses the review commissioned by the IAGCI from Dr John
Campbell (which he describes as being “unequivocally scathing" about the
September Home Office CIGs), without any reference to the Home Office response
to the IAGCI review published in November 2015. Pressed about this by Mr Rawat,
PK said he had read the Home Office response but saw no need to comment on it.
We find that a lapse in judgement on the part of the professor. Irrespective of
whether or not the Home Office response was cogent, any endorsement of the Campbell critique without reference to it was bound to appear as a one-sided treatment of
the relevant materials. It does not comport with proper performance by an
expert witness of his duty to identify evidence pointing against as well as for
the opinion evidence proffered.
232.
For
similar reasons we find it striking that despite continuing into 2015-2016 to
devote a significant portion of his critique of the DFFM Report to
highlighting the claim by Drs Olsen and Olesen that the head of mission had
put pressure on them he nowhere mentions that the Danish Ombudsman specifically
examined the question “Was pressure put on staff of the DIS to
paint a favourable picture of conditions in Eritrea
which were not actually how
things were?” and reached the specific finding that “I
have no reason to believe that the DIS wished to give the conclusions in the
report an untenable expression or put pressure on its staff with this purpose
in mind” (see above [181]). None of this is to gainsay whether in fact the
Ombudsman’s findings amounted to a direct contradiction of the evidence of
Olsen and Olesen or whether the remit of the Ombudsman was sufficiently broad
to enable him to make findings on such matters. But the Ombudsman is the only
person who had sight of all relevant documentation about what actually happened
inside the DIS regarding the DFFM and at the very least his findings on this
issue should have been referred to. PK’s failure to refer to them coupled with
his great emphasis on what Olsen and Olesen had alleged, was quite
insufficient. His reference to the Ombudsman’s findings otherwise do not
rectify this insufficiency.
233.
There
is also an element of one-sidedness in PK’s treatment of certain NGO sources
who became involved in the polemics surrounding the DFFM Report. In his report
for AA, for example, he sought to rebut the view (which he saw to lie behind
the DFFM Report) that only informants inside Eritrea
could produce reliable
data by invoking “highly reputable and dedicated human rights organisations,
such as AI, HRW, Journalists Without Borders etc. who have over time built
formidable reputation thorough unimpeachable rigour and scrupulous scrutiny and
seasoned academics whose publications are filtered through severe and thorough
scrutiny”. Under cross-examination on the subject of the AI report, “Just
Deserters”, PK had to agree that he was not in fact in a position to vouchsafe
the contents of that particular report apart from the fact that it comported
with his own understanding of the situation in
Eritrea
. The strong reputation
of bodies such as AI is not in dispute and is often referred to by various
courts and tribunals, including this Tribunal, as one relevant consideration
when assessing sources, but it is not in anyone’s interests, including AI’s,
for such respect be elevated into a dogma. As PK accepted in his oral testimony,
every report from whatever source must be subjected to the same critical
standards. When adulatory language like this co-exists with ad hominem
criticism of the motives of western diplomats in
Eritrea
, the inevitable
impression created is of an expert who has strayed from an approach that is
unwaveringly objective and impartial.
234. We would emphasise that we find these aspects of PK’s written evidence atypical and note that in his “Reflections on Home Office FFM”, 21 April 2016, he generally adopts a far more measured and objective tone (save for one isolated passage in 5.11).
235.
This
case has required us to examine many key sources under a microscope in relation
to the methodology underlying them and in this context it is almost inevitable
that PK’s own research techniques and methods should come under greater
scrutiny than ever before. To a certain extent we think that some of the criticisms
directed by the respondent at his written and oral evidence for defective
methodology were over exacting: e.g. it was suggested that there was something
deficient about the fact that he said he had only conducted 24 research interviews
since 2012 (between 2002 to 2012 he had conducted 190). Viewed together with
his daily contact with a network of sources and his various papers and talks on
Eritrea
, this does not strike us as deficient. The criticism that he failed
in his AA Report to cite the current Tribunal Practice Directions is not
unimportant because these contain the basic ground rules for an expert witness,
but the respondent does not dispute that the earlier version on which the
professor relied was in substance the same. On the other hand, we do consider
that the respondent identifies some significant shortcomings in his patterns of
research. We have already had cause to refer to his regrettable lapse in
sending an email describing the DFFM draft report as “well-written and
informed” when he had not even read it and to his evident inattention to the
meaning likely to be conveyed by passages he approved for the DFFM notes.
Despite the claim in his AA Report that “I have a practice of counter-checking
every source”, he conceded in cross-examination that this was not always the
case. There were instances, hopefully isolated, where he exhibited carelessness
in sourcing (e.g. using a book published in 2013 making an un-sourced claim
that the people’s militia had increased draft ages to 80 and failing to note
obvious “round tripping”; e.g. his description of Martin Plaut’s 26 February
2016 “analysis” when the latter was simply cutting and pasting from HRW).
236.
More
troubling to us in terms of determining the weight to be given to his expert
evidence has been the information he has given in response to questions seeking
clarification of who comprise what he referred to several times in his reports
as his “dense network” of sources in Eritrea
. Whilst we have no doubt that his
network is a significant one and includes contacts with persons who work in the
Eritrean government as well as outside it, we were surprised at the number of
occasions when under pressure from Mr Rawat, it turned out his information for
a particular matter comprised just friends and family and/or was based on an what
he described (too often) as “common knowledge”.
237.
It
was also not always easy to tell when he was drawing on his own research
interviews and/or his dense network of contacts and when he was rather relying
on his own unadorned opinions. Sometimes he was very candid about relying
solely on the latter, as when in reply to questions about the likely profile of
Eritreans who go back to Eritrea
for holidays, he said “but I’m speculating”,
yet prior to that observation he had given the impression his opinion on this
matter was based on his dense network of sources. Sometimes when pressed he
explained that he arrived at his opinion simply by inference or deduction from
other known facts. For example, he said that he had reached the conclusion that
the category of students able to obtain exit visas had narrowed because this
was a “consequence” of the (greater) numbers of Eritreans leaving the country
and the introduction of the people’s militia. We fail to see what added value
such comments bring to understanding of this issue. We did not always find it
easy to follow how he reached certain of his key conclusions: for example when
asked why he had not stated prior to 2016 that the upper age limits for exit
visas had changed from what he had stated them to be in MO, he said he
had changed his mind as a result of his research. Yet elsewhere he said he
relied for information about this on his own “dense network” of contacts in
Eritrea
. Given that the people’s militia was established in 2012, we do not understand why
it took four years for him to revise his view about this.
238. We were also troubled by his evidence relating to the paragraph in the DFFM which he agreed: see [36] and [175] above. Although we accept that the DIS interpreted it in a way that does not represent his position, PK’s evidence about his interpretation of it and what was meant by it was unimpressive. He sought to explain the identification of “a few deserters” by reference to three witnesses, but when probed about the three, it is clear that his information about them was lacking in detail. He described the “relaxation” he meant as being in the government’s attitude to its supporters, but this did not make much sense to us.
239.
In
light of such observations we find ourselves unable to attach as much weight to
PK’s evidence as the UT has done hitherto in country guidance cases on Eritrea
. We continue to view him as an able academic having a long-established and
extensive knowledge about conditions in
Eritrea
and someone whose research
plays (and we hope will continue to play) an invaluable role in informing
others about the nature of a regime which makes it particularly difficult for
the outside world to gain a full picture of what happens inside the country.
However, whilst for that reason we continue to draw on his evidence as one of
the many sources available to us on
Eritrea
, we are not able to give it
pre-eminent weight.
240. It is fair to add straightaway that our re-evaluation of PK as an expert witness has not in the end had a significant effect on our main conclusions, since we now have considerably more evidence from other sources, including of course the two UNCOI Reports of 2015 and 2016. Mr Knafler correctly observed in submissions that the appellants’ cases did not hinge on whether the Tribunal felt able to rely heavily on PK’s evidence.
D. FINDINGS ON MAIN GENERAL ISSUES
241. We are now in a position to give our findings on the main issues arising in this case. As signposted already, we consider it best as much as possible to set down in the same place first a short synopsis of the relevant background materials; second a reference to any existing country guidance on the issue; third, a brief outline of the submissions we had regarding each issue; and fourth our findings on it. We shall deviate a little from this structure where appropriate. It is in the nature of the main issues thrown up by the Eritrean context that some overlaps will occur.
The general situation
242.
It
is not in dispute that the human rights situation in Eritrea
remains of deep
concern.
243.
The
background reports chronicle some positive measures which seem to us to be
uncontentious. For example, in February 2016 a delegation of OHCHR was
permitted to make a working level technical assessment visit to Eritrea
. In addition a delegation of HCHR visited
Eritrea
in March 2016 and was permitted a
short visit to Sembel prison.
Eritrea
has acceded to the Convention Against
Torture (in September 2014). The government has brought into force a new Civil
Code, Penal Code, Code of Civil Procedure and Code of Criminal Procedure. It
has adopted Proclamation No. 158/2007 to abolish female genital mutilation. It
has made some progress in achieving the health-related Millennium Development
Goals. It has formulated a new national policy on children. In March 2015 the
European Commission under the
Eritrea
-European Union Partnership of 2015 and
National Indicative programme for
Eritrea
, 11th European Development
Fund announced a new development package of 312 million euros. In April 2014, 8
political detainees were released and in January 2015, 6 journalists were
released from detention. There is some evidence of a raising of national
service salaries, the printing of new currency rates to deter
people-traffickers and greater foreign investment in mining and other sectors.
Checkpoints for ID and travel documents are less prevalent. There appear to
have been more journalists able to visit Asmara and sometimes other areas.
There is evidence that many diaspora Eritreans return to
Eritrea
each year to visit family and friends.
244.
There
is, however, far more frequent mention in the background evidence of continuing
matters of concern. Those that are uncontentious include, for example, the fact
that the UN Security Council continues to extend the arms embargo on Eritrea
. The 1997 Constitution, published as the Supreme Law of the Land, remains
unimplemented and indeed the President declared the new Constitution void on 30
April 2014. The National Assembly remains suspended and there is an absence of
a functioning legislature. There is no independent judicial system. The 2015
World Press Freedom Index ranks
Eritrea
last among 180 countries. The economy
is weak. The 2015 UNCOI Report cites a former military clerk as stating that
one could estimate the number of detainees to have reached 14,000 in 2014 in
military prisons alone (see [794]). The same report states that torture is
widespread and systemic. The level of corruption increased such that in 2015
out of 168 countries only 13 others were ranked as more corrupt than
Eritrea
(see Transparency International 2015 Index).
245.
It
must be emphasised that as regards the debit side of the human rights auditing
of Eritrea
, the respondent’s position is little different from that of the
appellants and UNHCR, although she clearly maintains that on certain issues
relevant to risk categories on return there has been some improvement. This is
not surprising given that the FCO, for example, continues to issue statements
expressing very serious concerns about the human rights situation in
Eritrea
: see above at [85]. In outline submissions Ms Dubinsky identified ten propositions
which can be derived from the evidence which she understood not to be in
dispute between the parties. Although neither Mr Rawat nor Mr Knafler agreed
these expressly, we consider it provides useful context to set them out as
stated by Ms Dubinsky (it is only really as regards point 10 that there is any
obvious conflict).
1).
Eritrea
is a ‘closed state’ in which independent media have been banned since 2001 and
the government has resisted international human rights monitoring by the UN
Special Rapporteur for
Eritrea
, the African Commission of Human Rights and
established NGOs such as AI. Even the ICRC, which has a presence in the
country, is not given access to prisons. As regards its visit to Sembel prison,
HCHR stated that the visit was not carried out “…in conditions that allowed for
full human rights or technical assessment…No specific information was provided
on the number of detainees, not on their identity, safety, well-being or
whereabouts” and OHCHR said it “remained concerned about continued reports and
allegations of serious human rights abuses”. There are restrictions on the
ability of international diplomats and the representatives of international
organisations to travel outside Asmara although some travel takes place through
a permission process.
2).
Eritrea
is a one-party state. The only recognised party remains the ruling PFDJ and
there is no indication that provisions contained in the 1997 constitution which
would have allowed other parties to exist will be implemented or contained
within ongoing discussions for a new constitution.
3). There is a continued undeclared state of emergency which government representatives justify on the basis of the ‘no war, no peace’ policy and continuing concerns about Ethiopian belligerence. Given that there continue to be border skirmishes, it seems unlikely this policy will alter in the near future.
4).
Eritrea
operates both ‘official’ and unofficial detention sites, the latter which
include underground cells and shipping containers. Detention is often
unrecorded.
5). There is an absence of rule of law. As already noted, the judiciary is not independent, trials fail internationally recognised safeguards and detainees continue to be held for long periods without charge and incommunicado. There are no known internal or external mechanisms to investigate security force abuse.
6). Torture remains widespread.
7). Despite recent government indications that it would set an 18 months limit to national service, it has not done so and has disavowed its intention of doing so.
8). It remains a criminal offence in Eritrean law to exit the country illegally and to desert or evade national service.
9). MA and MO were correctly decided at the time.
10).
With
very limited exceptions Eritreans between the ages of 5 and at least 54 (men)
and 47 (women) are prohibited from leaving Eritrea
.
246.
However,
despite considerable common ground between the parties over country conditions
in Eritrea
, they disagree over whether there have been significant improvements
in certain respects that bear on the issue of risk on return. That being so,
the extent to which
Eritrea
’s very poor human rights record in general informs
our assessment must depend on an issue-by-issue analysis.
National Service
247.
According
to the EASO Report, May 2015, “Eritrea
’s national service ... differs from the
defence forces of other countries in that its overall aim is not only to defend
the country, but also to rebuild it following the War of Independence and to
propagate the relevant ideology”. National service is regarded as “the school
of the nation”.
248.
The
same report records that the CIA World Factbook estimates Eritrea
’s population
as of July 2014 at just over 6.3 million and that the manpower reaching
militarily significant age annually is around 66,800 males and 66,700 females.
According to the 2015 UNCOI Report at [1178], there are an estimated 201,750
civilian active members of the armed forces who are national service conscripts.
Under the Proclamation of National Service No.82 (1995 [
Eritrea
]) Article 8,
“Active National Service” consists of six months of training in the National
Service Training Centres and twelve months of active military service and
development tasks in military forces. As a result of the ‘no war, no peace’
policy
Eritrea
adopted following the war between
Eritrea
and Ethiopia (1998-2000), the government launched the ‘Warsai Yikealo Development Campaign’ (WYDC).
Thereafter national service was considered as indefinite. Eritrean law
contains no provisions for conscientious objection or alternative service.
249.
Article
22(1) of the 1995 Proclamation states that “[t]he citizen who upon termination
of military training enters into a 12 months of Active National Service is entitled
to pocket money”. In 2015 this was said to be less than £6 per month ($10– see
‘African Dictatorship Fuels Migrant Crisis:
Thousands flee isolated Eritrea
to escape life of conscription and poverty, M.Stevis
and J.Parkinson Wall Street Journal, 21 October 2015).
There is evidence before us of government promises to increase this amount and
of this having been done in some cases for some periods, but it is far from
clear that these promises have been implemented on a general scale.
250. The Proclamation of National Service specifies, as categories of people who may be exempted from or unable to perform military service:
“Those exempt from ‘Active National Service’ are (1) the citizens who have performed national service before the Proclamation; (2) all fighters and armed peasants who have spent all their time in the liberation struggle (Article 12); citizens who suffer from disability such as invalidity, blindness, psychological derangement” (Article 15(1)).
251. As regards students, it is stated in Article 14 that:
“Students on a regular daily course may be exempted from Active National Service for a limited period (a) if he is continuing his studies from middle up to secondary grade; (b) if he is following his course of studies in a Professional or Technical School; (c) if after passing university examinations he has been accepted by the university and is following his studies; (d) if he has been authorised as a special case to continue higher studies by the Technical School or by the university; (e) if at any school level he has been required by the government to attend a special course or to be sent on a scholarship.”
252. Article 15(2) states that ‘[t]he citizens who [....] are declared exempted from national service by the Board will receive from the Ministry of Defence a certificate of exemption’.
253. Prior to the National Service Proclamation of 1995, married women and mothers were exempt from national service. According to PK and the 2015 UNCOI Report, although the 1995 Proclamation removed these exemptions de jure for married women and mothers, many married women and single mothers continued to be de facto exempted at the discretion of recruiting officers (2015 UNCOI Report at [1201]).
254. Eritrean national service can involve civilian service which can include jobs in the civil service. Indeed the great majority of conscripts are engaged in civilian national service rather than military national service.
255. The aforementioned Proclamation sets out a penalty of 2-5 years’ imprisonment for military violations.
256.
All
12th grade students, including some younger than 18, are required to
complete their final year of education at the Sawa Military and Educational
Camp. Those who refuse to attend cannot receive high school graduation
certificates, go on to higher education or be offered some types of jobs.
Anyone who drops out of school before their 11th school year can be
conscripted for national service directly by the Kebadi Administration once
they reach the age of 18. In 2014 the government announced that
the duration of national service of future conscripts would be limited to 18
months. It was PK’s unchallenged evidence that on 25 February 2016 the
Eritrean Information Minister, former spokesperson of the president, announced
that there were no plans to scrap or cut national service. PK’s evidence
throughout is very firmly that national service is indefinite and he referred
to members of his family who had served 16-17 years and 20 years
respectively. The EASO Report at [3.7.1] referred to a study by PK in 2008 and
2012 among Eritrean migrants in European and African countries which revealed
an average service time of 5.8 years. PK’s evidence is that this was taken out
of context and what was meant is that this was the average time of service
prior to fleeing Eritrea
. His evidence is that it is open ended.
257. AI (the AI Report on AA at [72]) states that it is commonly accepted that Eritrean national service is indefinite in duration both for those engaged in military and non-military activities and what this means is that it is of unknown duration rather than permanent and it is subject to arbitrary and unpredictable recall. Interviewees told AI that they have been in national service for 7 and 8 years respectively. AI notes in “Just Deserters” (at page 15) speaking to a woman whose husband had served 20 years and another whose husband had been in national service since 2006. The UKFFM interviewed anonymous witnesses who claimed that many had done it for 10 – 20 years.
258. The 2015 UNCOI Report at [206] notes that numerous witnesses gave an account that the duration of military service is arbitrary and often of punishing length and routinely outside of the eighteen month period provided for in the 1995 Decree. It is frequently for periods well over a decade. The Eritrean government has repeatedly justified the prolongation of national service with what they consider to be the continued occupation of its sovereign territories and the so called “no war, no peace” situation. Conscription into the national service is at an early age without any prospect of being formally discharged or otherwise released (see [1250]). The Commission of Inquiry interviewed witnesses who had been in national service for periods including 17, 18 and 14 years. The procedure for release is “unclear” (see 1252]), there being no rules or procedure governing this or mechanism to challenge a refusal. Journalists report people being in national service for more than a decade (Mary Harper and Edmund Blair).
259. On the other hand there are a number of reports which, whilst recognising that national service is indefinite and that duration can be lengthy, give analyses that suggest or lend some support to the view that on average most Eritreans have completed their national service before the end of 7 years. We will address this evidence below at [304] – [307].
People’s Militia
260. In 2012 the government created a new programme called Hizbawi Serawit or the people’s army or militia. According to the 2015 UNCOI Report at [144] the motivation behind its introduction was “perhaps in response to an increasing number of defections, dwindling numbers of conscripts and ongoing incidents with neighbouring countries”. This is described as a compulsory system providing for additional military training as well as assignment to unpaid law enforcement and other civilian duties, such as agricultural work, development projects and security and border guard duties. At [201] of the 2016 UNCOI Report the Commission stated that it had “received numerous corroborated reports that Eritreans in their 60s and 70s have been forced to participate in the people’s army, as well as persons who had been released from military/national service on health grounds.”
261. People’s army units are said by the 2015 UNCOI Report at [286] to be organised by profession (e.g. teachers’ militia, artists’ militia) or by geographic area or neighbourhood. Units meet regularly, i.e. one day per week or one week per month. Members are allowed to keep their current jobs (see [286]). There is no known law or decree regulating this programme.
262. As such the people’s militia constitutes a form of compulsory service. Although separate from national service, it now constitutes part of the Eritrean military service system.
263. Given the dearth of clear evidence as to age-limits, we consider we should regard the age limits as being that contended for by the appellants and UNHCR, namely up to 70 for men and 60 for women.
264.
As
regards the nature of military/national service in Eritrea
, the Tribunal in MO
did not seek to make any finding on it except insofar as it was relevant to the
“issue of categories of lawful exit and risk on return for those who had left
illegally” (see [3]). It reconfirmed the findings of the Tribunal in MA
that national service in
Eritrea
is open-ended and indefinite and
demobilisations from active military service did not free people of ongoing
obligations to undertake other types of national service. As explained earlier
in the section dealing with the country guidance issues addressed in this case,
we have accepted that there is now a need to address national service issues
more directly, including the issue of whether military/national service in
Eritrea
is contrary to human rights prohibitions on slavery, servitude and forced or
compulsory labour. Potentially, for the appellants and UNHCR, such an
assessment could result in a conclusion that irrespective of whether a person
left
Eritrea
illegally or not, as long as it was reasonably likely they would
be required to perform national service, they would on return face a real risk
of being exposed to treatment contrary to Article 4 of the ECHR. This is
considered separately below.
Submissions
265.
The
respondent’s position is that no previous country guidance case has reached the
conclusion that the prospect of an individual undertaking national service is
in itself sufficient to prevent him or her from lawfully being returned to Eritrea
. The issue had only been relevant insofar as a person will be regarded as a
deserter or draft evader and subjected to punishment.
266. The respondent submits that the position in 2016 discloses a far more nuanced position than confronted the Tribunal in MO. She considers that there are a significant number of people who appear able to obtain an exit visa, which chimes with other evidence that there are viable exemption categories and that there are a significant number of demobilised persons who are able to obtain exit visas. It is, she submits, a proper inference from the evidence that a large percentage of the Eritrean population is not involved in national service. At most, the numbers engaged in national service are only 9 per cent of the population. This supports the argument that the Eritrean state does not assiduously pursue those who have not done national service.
267. The respondent accepts that the Eritrean government had not gone ahead with its promises made in 2014-2015 to limit the duration of national service to 18 months. Her position is that its duration is variable and to a degree uncertain, but inherent in the variation must be the real prospect of being discharged from national service. The respondent does not accept that only a few have been discharged or that discharges are only available on the grounds of ill-health. The criteria may be inconsistent, but the following categories can be identified: those who are discharged simply due to the passage of time or on request; women; those suffering from physical or mental health problems; those with contacts or who are able to pay bribes; and those seeking discharge for economic or family reasons.
268. As regards the conditions of national service, the respondent argues that there is a very significant distinction between military service and those carrying out national service in civilian roles. That was recognised by the 2015 UNCOI Report at [1443] and also by other sources. This distinction also impacts on potential punishment. The same UNCOI Report states at [1446] that conscripts in civil service are usually not subjected to harsh punishment.
269. As regards punishment for draft evaders and desertion generally, the basis of the respondent’s position is that the latest evidence indicates that the Eritrean government no longer detains them routinely or exposes those they do detain to physical harm. Punishment taking the form of ill-treatment is reserved for people who have had some sort of oppositional activity or where for symbolic reasons the government wants to make an example. The “shoot to kill” policy and the round-ups of suspected evaders (giffas) and the targeting of relatives are now, she submits, significantly less likely occurrences.
270. The respondent disagrees flatly with the appellants’ contention that military service, and the people’s militia all amount to violations of Article 4 of the ECHR. The evidence does not indicate, she submits, that there is a consistent pattern of such violations, especially given the evidence that there is a wide spectrum of circumstances and that the national service system is variable as between military and civilian service and between national service and the people’s militia. For many individuals national service amounts to no more than attendance at an office part-time or in working hours, in Asmara, living with their families. Service in the national service is variable and might amount to duties once every two weeks. Demobilisation or discharge can be anticipated as a likely outcome for many. As regards conditions, the 2015 UNCOI Report itself states at [1426] that “[t]he length and the conditions of work for conscripts, including wages, working hours, places of assignment, leave time and rest days, do not per se constitute elements of forced labour”.
271. The appellants’ position as regards national service is distinctly at odds with that of the respondent. According to their submissions the risk categories identified in MA and MO require expanding so as to state that all citizens between the ages of five and 70 (regardless of gender) who are not medically unfit for national service and who have left illegally are viewed as being at risk of serious risk of persecution, save for limited categories. It was wrong of the respondent to suggest that many Eritreans did not go to national service. National service remained open-ended with no meaningful demobilisation and with eligibility extended to those between the ages of 60-70 years in the people’s militia. Conditions in military service remain oppressive, very harsh and life-threatening and amount to a breach of Article 4 of the ECHR. They rely on the 2015 UNCOI statement at [1501] that “[t]here is a pattern of torture, inhuman, cruel or degrading treatment or punishment of conscripts in the army in connection with the labour that conscripts are forced to perform”; and the ILO Observation (CEACR) which accepted in 2015 that “Eritreans are subject to systems of [national service] and forced labour that effectively abuse, exploit and enslave them for indefinite periods of time.....”; and the USSD Report of 13 April 2016 which referred to examination by the International Labor Organization Conference Committee on the Application of Standards noted discussion “relating to the large-scale and systematic practice of imposing compulsory labor on the population for an indefinite period of time within the framework of the national service program which encompassed all areas of civilian life and was therefore much broader than military service.” Punishments often amount to ill-treatment.
272. In relation to the treatment of those returnees who had left illegally, the appellants’ submission is that it was extremely unlikely that the Eritrean government had changed its position, not least because to treat them leniently would place them in a more advantageous position than those who had not left illegally. The letter of regret does not establish that they are excused their obligation to perform national service in the future and effectively amounts to a confession of guilt.
273. The appellants submit there was no evidence of significant demobilisation or discharges and that there is no provision for exemption from national service to women and the practice of exemption in medical cases was arbitrary and difficult; there is no specific sole breadwinner exemption.
274. The appellants submit that whether one is conscripted into the military or civil service within the national service framework makes no difference because no conscripts can leave the country and none are free to change employment. Conscripts in civil service are subjected to the same restrictions on movement as in the army and travel permits are limited to their area of service. Conditions may be variable but fundamentally the consequences are the same.
275. For UNHCR there has been no fundamental, durable or stable change since the MA and MO country guidance. National service remains open-ended and can last more than ten years in practice; its conditions are harsh, amounting to inhuman and degrading treatment. Allocations to different forms of national service and military exemptions are arbitrary as is the duration of national service; and individuals may be transferred between different types of national service assignment as a punishment. There is therefore a real risk of re-conscription to a harsh military assignment even for a person previously given an informal exemption or previously allocated to relatively light civilian duties. Women and girls face the additional and real risk of sexual abuse. Conscripts are reportedly paid extremely low wages and are routinely separated from their families. They are forced to carry out non-military work, such as construction, mining and farming in circumstances amounting to forced labour. Punishment for transgressors, including evasion and desertion is draconian; exemptions are limited and even the legal exemption on medical grounds is applied inconsistently and granted only in exceptional circumstances.
276. In relation to round-ups and the “shoot to kill” policy, the position of the appellants and the UNHCR is that these are still part of the regime’s policy and that family members of evaders and deserters are also punished. The evidence they rely on includes the AI statement that round-ups take place regularly (see the AI Report on AA at [93]) during which anyone appearing to be of national service age who cannot produce papers justifying their absence from service is taken into custody. According to interviewees spoken to by AI, the round-up is a “regular threat” and at [95] it is stated that round-ups occur at schools. It is asserted at [96] that some who produce exemption documents are rounded-up in any event and ordered to return to service. The 2015 UNCOI Report at [1211] reports that the Eritrean Defence Forces regularly conduct round-ups in search of citizens who have failed to respond to national service call ups or who have absented themselves from the army without leave. Excessive force is often used (see [1229]). Anonymous source 1 interviewed by the UKFFM said that round-ups were random and that there was one three or four months ago in Asmara whilst anonymous source 3 stated that a round-up occurred only two weeks before.
277. AI reports (in the AI Report on AA at [132]) that anyone caught illegally crossing the border is arrested and punished and that the “shoot to kill policy” remains in force for those crossing the Ethiopian border and there were shootings. It is asserted by AI (“Just Deserters” page 52) that the policy is only in force on the border with Ethiopia. An incident in 2014 is cited. AI spoke to conscripts who were deployed on the border and one, who left in 2015, stated that they let people get across the border but had they been caught by their commanders they would have been in trouble and even killed and that sometimes they would obey orders and shoot people. Those caught trying to flee face detention for anything from a few months to many years. Those assisting others to flee the country face longer periods of detention. There were reports of torture during interrogation and general mistreatment.
278. The 2016 UNCOI reports at [56] that the “shoot to kill policy” has not been rescinded, but that it has been implemented in a less rigorous manner in recent years. It is reported in the 2015 UNCOI Report (at [1234]) that desertion can be punished with the death penalty and at [1115] and [1116] it is concluded that the number of shootings and killings is high and reference is made to incidents in 2013 and 2014. The Commission of Inquiry also heard from witnesses that there had been a revision of the policy and that border guards are under orders to shoot below the knee with a view to stopping the flight after firing a warning shot in the air, but there was evidence from others interviewed that border guards still continue to shoot at people who attempt to cross the border whilst others said that they had crossed the border without problems as border guards no longer shoot at people. The EASO Report cites an incident in 2014 (see 6.4.3). PK‘s evidence was that the policy was in force citing a shooting that took place in Asmara on 3 April 2016.
279. The 2015 UNCOI Report describes reprisals against family members, friends and associates following the conduct of a third person which include arrest and detention. The Commission records the accounts of witnesses who had fled that detail family members having been detained and arrested. Landinfo in the report of 23 March 2016 reported that since 1999 threats have been made against family members of deserter/evaders. PK’s evidence was that family members were targeted, but that his own family had not been, despite his open criticism of the regime and overall he thought the policy was less in evidence. Anonymous source 2 spoken to by the UKFFM heard of reprisals against families of those who fled illegally including the detaining of a parent until the child’s return, but that it did not happen much at present because of lack of manpower.
Our assessment
Enforcement and punishment
280.
Before
addressing conditions, we shall first of all address punishments because it
seems to us that notwithstanding the respondent’s submission that the Eritrean
authorities have adopted a “more pragmatic approach” as stated in submissions
and in the Country Information and Guidance: Eritrea
: National
(incl.Military) Service, Version 3.0, August 2016),
the preponderance of the evidence points strongly to the conclusion that the
Eritrean regime of military/national service (excluding civilian national
service and the people’s militia), is characterised by a system that often
responds to transgressors with harsh and disproportionate punishments. We
exclude from this conclusion civilian national service and the people’s militia
because by contrast the evidence does not demonstrate that punishment for
transgressions by persons evading or deserting from one or the other is either
as likely or as severe in nature.
281. We would accept that the preponderance of evidence also indicates that roundups (giffas) are happening less frequently and that the “shoot to kill” policy is now intermittent and arbitrarily applied and that punishment of family members or associates may not be as common as it was, but these are only some of the regime’s repertoire of punishments, and there is a substantial body of evidence, including the US State Department reports, indicating that the generality of evaders and deserters are harshly punished and this is a common thread running through the majority of source evidence. We note that the 2015 UNCOI Report at [818] refers to the grant of an amnesty to deserters in November 2014, but this was from detention and the Report does not suggest this represented a change of government policy. The main evidence concerning this matter on which the respondent relies is that from Eritrean government ministers and interviews with individuals during the UKFFM and we have explained why we feel that this evidence should be approached with caution: see [192]-[201]. We have taken into account the evidence of AI (the AI Report on AA at [104]) that punishment for deserters is generally more severe although this is arbitrary and that the generality of evaders and deserters are punished with imprisonment for varying periods. Those caught on the border trying to flee are almost always subjected to periods of arbitrary detention. Generally (see [106]) those arrested for evading service are detained for some time between one and six months. The reports demonstrate (see [115]) a high level of variation which is said to be indicative of the arbitrary nature of punishments that are at the discretion of officers. The EASO Report concludes at [3.8] that deserters and evaders are punished by imprisonment if caught within the country before being able to leave or on return at the airport and that punishment is harsh being more severe for deserters. PK’s evidence throughout is that deserters/evaders will be subject to persecution.
282.
The
2015 UNCOI Report (at [96] and [97]) reports arbitrary detention, enforced
disappearance, torture and mistreatment generally in Eritrean detention
centres. The Commission spoke to those who had fled in the past two years and
reported that they had been subject to ill-treatment and detained without due
process. The Commission (at [239]) reports arbitrary detention for periods
ranging from months to years, enforced disappearances ([249]) and torture
([259]). EASO reports (at [4]) poor conditions in detention. The Swiss fact-finding
report of March 2016, to which several references are made in the new Home Office
CIGs of August 2016, considers that even though the treatment of deserters
appears to have become less harsh in recent years, “[m]ost sources report that
first time offenders are now usually detained for several months” (cited in the
CIG on National Service at 15.2.18). Given that we consider anything beyond
very short-term detention in Eritrea
to create a real risk of ill-treatment,
this confirms our view that deserters/evaders continue to face a real risk of
persecution.
283. To summarise, we reject the respondent’s case that enforcement and punishment is reserved for those who are involved in oppositional activity over and above desertion or evasion. It is impossible in our view to derive from the evidence as a whole any other conclusion than that for Eritreans inside the country any evasion of military service or desertion still carries a real risk that the generality of transgressors will be subject to treatment which amounts to persecution as well as serious harm.
Conditions
284. As regards conditions of national service, we will have cause to return to this subject when considering the interrelated issue that has arisen in this case as to whether the Eritrean system of national service is in violation of Article 4 of the ECHR, but we need here to give our general view on conditions in broad terms. In light of the finding we go on to make, that those who left illegally and who would be perceived on return as draft evaders/deserters would face a real risk of persecution and serious harm, it will not be relevant in most cases to consider whether a forced returnee would be at real risk of facing national service that was in breach of their human rights (by virtue of the system of military/national service being abusive because of conditions or other features). However, we recognise that there may be cases of persons facing forcible return even though they are not perceived as draft evaders or deserters, for example because they exited lawfully. Hence it is necessary to make specific findings on the system of military/national service, including as regards conditions.
285.
We
recognise that hitherto Tribunal country guidance on Eritrea
has not regarded
the Eritrean system of military/national service as generally demonstrating a
consistent pattern of breaches of human rights. We continue to be cautious
about making distinct findings on this issue. One reason is that the principal
source relied upon by the appellants and UNHCR – the 2015 UNCOI Report – does not
make sufficiently clear findings about the conditions in Eritrean national
service today, as distinct from the conditions pertaining during the past three
decades. If its position is that for all points in time over this period (including
the present) conditions demonstrate such a consistent pattern, then we would
expect greater clarity about this. The 2016 UNCOI Report, although having a
remit to consider the last 12 months, refers often to examples from previous
years. Another reason for caution is that it is quite difficult very often to
disentangle what is stated in the UNCOI reports and the other background
evidence about conditions and what is stated about the punishment regime for
those who commit transgressions. As we have just said, we are in no doubt that
the latter is abusive, but evidence about that does not necessarily establish
that conditions as such are abusive. Another reason to be cautious is that the
UNCOI reports themselves identify (as do the sources cited in the Country
Information and Guidance:
Eritrea
: National (incl.Military) Service, Version
3.0, August 2016, to mention but one recent compilation) that
conditions in civilian military service (and in the people’s militia) are
better relatively speaking.
286. Nevertheless we consider that the evidence for finding conditions of military national service (not civilian national service) generally abusive is stronger than was the case when MA and MO were decided. Despite the reasons expressed above for being cautious regarding the UNCOI Reports, we regard what is said at [1391] of the 2015 Report as broadly reflective of the wider body of evidence, namely that “conditions of national service characterised by conscripts’ lack of adequate food, access to water, access to hygiene facilities and adequate accommodation during military training and service, constitute cruel, inhuman or degrading treatment” ([1391]). (Why we consider that although conditions are not generally abusive for those doing civilian national service the national service system generally constitutes forced labour contrary to Article 4(2), is dealt with in our section below on Article 4 of the ECHR: see [376]-[430]).
Eligibility/duration
287. As regards the eligibility requirements for national service, age (and duration) in particular, we will deal below with the age requirements when considering the categories of lawful exit visas: see [308]-[328]. But in a nutshell we consider that the age limits for national service are likely to remain the same as stated in MO, namely 54 for men and 47 for women except that for children the limit is now likely to be 5 save for adolescents in the context of family reunification. For the people’s militia, the age limits are likely to be 60 for women and 70 for men.
288. In relation to duration, it is agreed on all sides that national service is indefinite and open-ended, but there is disagreement as to whether this means that it results in most Eritreans performing military/national service duties permanently or for very prolonged periods. As noted above, the respondent’s position is that actual performance of military/national service is variable and uncertain, but that there is a real prospect of discharge. This is in stark contrast to the position of the appellants, UNHCR and PK.
289.
We
accept that there are no clear statistics relating to the number of individuals
in national service, but it is reasonable to infer from what evidence there is,
that at any one time most people are not engaged in the performance of military
or national service duties. Most sources estimate Eritrea
to have a population
of over 6.3 million. The 2015 UNCOI Report states that there are 201,750 active
members of the armed forces, the majority being national service conscripts (at
[1178]). The EASO Report at [3.1] states that there is no official data
available regarding the number of people engaged in national service but
various estimates place the figure at between 200,000 and 600,000 in recent
years, approximately half of whom are assigned to active military service.
PK’s opinion is that 9.2 per cent of the population has been conscripted over
the past 20 years (the figures, he states, do not take into account those who
have fled the country). The respondent does not accept PK’s percentage figure,
claiming that it is far less. However, even if we accept PK’s opinion on the
issue, which is the most favourable to the appellants, the figures are
significant. The only logical conclusion we can draw from them is that active
performance of national service duties cannot be as extensive as the appellants
and UNHCR assert. The system remains indefinite and
open-ended in the sense that all persons of or approaching eligible draft age
or within the age limits for the people’s militia remain obliged to perform military/national
service; but it is a distinct matter whether persons have to actually perform
military/national service and for what periods of time. We shall return to the possible
implications of this conclusion when we deal further with demobilisations/discharges
at [297]—[307].
290. We also consider the evidence to indicate that discharge/release is a more common phenomenon than the appellants contend. We will address this issue more fully when we deal with demobilisations/discharges and with draft evaders and deserters: see [297]-[307] and [338]-[356].
Exemptions
291. As regards exemptions from national service, we first consider the position of women. It seems to us that the source that is most representative of the various strands of evidence on this issue is the 2015 UNCOI Report.
292. According to the 2015 UNCOI Report [395]-[398], there is a “practice of tolerance with regard to women’s national service obligation when they are married or have children”. However, very few women have been formally released or discharged which makes it difficult for them to get identity cards or travel permits, although married women can get travel permits issued at an officer’s discretion.
293. As regards exemption on medical grounds, the evidence is mixed. There are a significant number of sources which state that physically disabled people are exempt from service. However, illustrative of a different view, in March 2015 Landinfo stated that people who are considered to be not fit for service are exempt from military training but must perform civilian service.
294.
We
note that there is wide recognition that (separate from the legal possibilities
for exemption, which all agree are limited by legislation to medical cases), a
significant number of people appear able to obtain exemptions based on contacts
and/or bribes. We take the principal thrust of the evidence regarding such
avenues as being that national service is not necessarily an unavoidable
experience for everyone in Eritrea
.
295. However, we need to say more at this point about medical exemption. AI in its Report of 22 September 2015 made observations in relation to the appellant AA, but also general observations that there is no functioning and reliable process of assessing medical fitness for national service. The Sawa training facility did not have a routine health assessment on arrival or at any time. The same is true for other camps. Permission to see a doctor or a designated first aid officer must be granted by a commanding officer and is extremely difficult to obtain. If assessed, resources are very limited. Recognition of mental health difficulties is harder to obtain than physical health problems. In any event, such exemptions are sometimes ignored in round-ups or call-ups. If conscripted it would be for an indefinite period and constitute forced labour. The UNCOI 2015 Report states at [60] that achieving exemption from national service is very difficult particularly for men. Examples are cited within the report of witnesses (with physical injuries) who had not been exempted and had been forced to remain in military service despite having been declared unfit (see [1196]). The Commission concludes that the exemptions on health grounds are rarely granted, even though the state of health of the persons concerned prevents them from serving in the military. There is evidence of blind and seriously visually impaired people being sent to Sawa ([see 1197]). In the UNCOI 2016 Report there is an example given at [92] of a witness who in 2014 was unwell with papers to establish this, but who was not believed. The witness reported being detained for six months without due process. AI in the “Just Deserters” Report of December 2015 refers at page 28 to former conscripts who told of people with disabilities being conscripted and taken to Sawa for military training. There is no health check or assessment of physical or mental fitness when people are first conscripted and sent for training or at the end of the year at Sawa. Medical assessments are carried out on an ad hoc basis, and usually only if the conscript repeatedly requests it. To obtain an exemption a doctor has to recommend that the conscript is unfit to serve, whether for physical or mental health reasons and this recommendation has to be confirmed by a military commander. Those with health problems have been assigned to national service and the report makes reference to a former conscript with a (physical) health problem who spent three years in national service and another former conscript who had severe injuries to both legs following a car accident. Although the commander concluded that he could not carry out physically demanding tasks, it was decided that he could work. The source stated that this is not a medical decision, but a decision of the commander. The individual was assigned to administrative work, but he was told there was no pain relief and not granted permission to see a doctor.
296. While we accept that there is an official exemption on grounds of ill-health, the majority of the evidence points to this being applied in an arbitrary and inconsistent manner. Whether a medical assessment takes place is not a given, but rather is entirely arbitrary. If a medical assessment takes place, the thrust of the evidence establishes that it would take place at the military training centre, rather than at a hospital. The outcome of the medical assessment is not determinative of exemption. The ultimate decision as regards exemption is not made by a doctor, but by a military commander. Relevant to this, there is on the evidence before us, a culture of disbelief on the part of the Eritrean authorities. It is not uncommon for those with mental health problems or general health problems to have to undergo national service.
Demobilisations/discharges and release
297. As regards demobilisations and dismissals/discharges, notwithstanding PK’s continued animadversions to the contrary, we consider the evidence taken as a whole establishes that these are now more frequent than was the case when MO was decided.
298. It is important first of all to express caution about the various terms that are used in the background literature. We find helpful the point highlighted in the EASO Report:
“A distinction should be made between demobilisations and dismissals; demobilisations follow wartime demobilisations; and dismissals take place on an individual basis after the discharge of obligations.”
but observe that there continues to be a lack of clarity both as regards terminology (dismissals and discharges and releases – and sometimes demobilisations - mostly being used as synonyms) and the meaning attached to various terms.
299. It seems to us also that even if achieved by unofficial means the Eritrean system of internal control on the movement of its population depends very much on persons carrying documents to prove their status and these can take various forms. They certainly include certificates of completion of national service; such documents also match the reference in Eritrean law to such certificates. We know from the 2015 UNCOI Report at [1254] that documents carried also include “release papers” which are obtained from employers. Although they are a precondition to a certificate of completion, they also serve as a travel permit.
300. We recognise that the 2015 UNCOI Report states at [1252] that the procedure for discharge/release is “unclear” and that testimonies reveal arbitrariness, systematic refusals of requests and no mechanism for challenge. It is asserted that release is rare and difficult to obtain and can usually only be obtained through bribery or for medical reasons and not on the basis of the number of years in service and that a certificate of completion is extremely difficult to get without facilitation by a high ranking employee of the Ministry of Defence. At [1261] it is stated that “reportedly, even persons who have documentary evidence that they have completed their active military service find themselves at risk of punishment as evaders of reserve responsibilities if they leave the country while still of military age”. But we consider that if discharge/release were not commonplace, the figures for those engaged in national service (even taking the highest given) would be much, much higher than they are. That release is a regular occurrence is the only rational explanation for the significant discrepancy between the appellants’ case and the numbers said to be in national service. It is also said that some holders of a certificate of completion have been recalled. For example, the 2011 UNHCR Eligibility Guidelines considered that people were liable to be recalled. The AI “Just Deserters” Report noted at page 26 a “recent practice of re-mobilising women or women and children in Gash Barka region”. Viewed as a whole, however, the evidence falls well short of indicating that this is a widespread phenomenon. In relation to national service, for instance, examples are given of people having been recalled when the war with Ethiopia broke out in 1998 and 1999 and remaining in national service. Anonymous source 2 spoken to by the UKFFM knew of someone who left national service unofficially and was working elsewhere, but who could be called back at any time. At [102] of the AI Report on AA it is stated that …”it seems that there may be some incidence of record keeping, of who has performed [national service], for how long and under what circumstances they left the country”. At [74] of the same report AI asserts that all those engaged in national service “…. including those at risk of round-ups and recall” are at risk of indefinite forced labour.
301.
We
find more compelling the evidence of the respondent, reinforced by the source
compilation contained in Country Information and Guidance: Eritrea
: National (incl.Military) Service, Version 3.0, August 2016, that discharge or
release is likely to be commonplace. The DFFM records Western embassy B’s
statement that “there are indications that young people are now released from
national service after a shorter period of service than was previously the
case” and Western embassy D stating that “[r]ecently, it seems …more and more
are released from national service after serving a shorter period of time”. We
note that in the context of round-ups, it would appear that many people who are
checked are able to establish that they are not evaders and deserters. For
example AI (“Just Deserters” at page 24) in the context of “giffas” gives
witness accounts of people with papers being released following a round-up. It
is not suggested that such round-ups result in whole communities being taken
off. We note also that whilst release can properly be described as arbitrary,
in that it is at the whim of a commander or employer and often on the payment
of a bribe, there is considerable evidence indicating that bribery and
corruption are prevalent. This seems to us to be borne out by
the language adopted in the 2016 UNCOI Report which observes at [159] that
“witnesses consistently linked corruption to exemption or early release from
military service”. They cite one witness as saying that “release from national
service is mostly by corruption”. We also know that a
substantial number of women are able, whether through family connections or
bribery or other means, to obtain a de facto discharge from military
service based on pregnancy, marriage, and responsibilities for children. The EASO
Report at 3.2 cites the Bozzini 2012 study’s observation that women over the
age of 27 can ‘regularise’ their status, i.e. be officially demobilised. The
Lifos Report of March 2015 at 2.9.2. and 4.2 notes women’s ability to get
demobilisation papers.
302. We re-emphasise that evidence to show that recall is a frequent phenomenon is lacking.
303. In reaching the above conclusions we have taken into account, inter alia, PK’s evidence on the issue which was in our view equivocal and unimpressive. On the one hand he said in relation to the evidence in the 2015 UNCOI Report about a certificate of completion that it was “inconsistent with [his] own analysis,” but then he said that some may be able to get papers, but added that he did not think that centralised records existed. He further explained that this was why there are round-ups because there are no centralised records, but this does not tally with the evidence cited above relating to round-ups and the showing of papers resulting in release. However, we accept that for the Eritrean government none of these discharges or de facto demobilisations means that their beneficiaries are thereby free of all liability to perform national service for ever more, except when they are outside the relevant age limits or perhaps in long-term medical cases. We also accept that because a significant number of discharges/releases appear to be de facto, it is likely that it will be difficult for beneficiaries to establish by documentation that they have completed national service. It is necessary also to look more closely at what is known about age limits and eligibility for national service.
304. A person starts national service at age 18 or indeed even younger in some cases. It is very unlikely that a conscript will be released within the first 18 months of service when a conscript is engaged in active national service (which comprises six months military training and 12 months military service). Our understanding is that immediately after this period conscripts are redeployed. The evidence points strongly, therefore, to a system which conscripts young people at 18 (or earlier) and then requires them to continue national service uninterrupted beyond completion of the initial 18 months. When a person starts national service, the term they will be expected to complete is not known and to this extent it is arbitrary and indefinite. Ordinarily, by the time they are in their mid-20s (unless they have been discharged or dismissed or released) they are likely to have been in national service for 7 years. The critical issue is how long the period is likely to be for them to be accepted to have completed national service in the eyes of the Eritrean authorities. Here there is evidence going both ways.
305.
On
the one hand, there is the evidence we have just noted that release is commonplace
and that for most citizens the duration is likely to be only several years. The
DFFM Report records Western embassy C as stating that it “had heard of people
in their forties who were still in national service, but in general 3-4 years
of national service seemed to be the norm” and Western embassy D as narrating
that “[t]oday it is easier to be released from the service and for young people
today national service seems to be limited to a couple of years”. The EASO Report
at 3.7.1. refers to two studies of Eritrean migrants where the persons were
conscripted for an average of 5 and 5.8 years respectively. The September 2015 Home
Office CIGs consider that the most up-to-date information available from inside
Eritrea
suggests in general that military/[national service] lasts for around
four years (a statement not seemingly retracted in the August 2016 version). The
UKFFM mission materials contain examples of persons whose national service was
relatively short. On the other hand, however, the evidence cited above is not
without problems. In particular one of the two studies cited by the EASO Report
is by PK and in his April 2016 Report he has pointed out, accurately in our
view, that the figures he gave in his study of 5 and 5.8 years were the average
years the conscripts interviewed for the study served before they fled the
country, not the years they had taken to complete their national service. The
other report mentioned by EASO, the “SIHA, Letters from
Eritrea
, Refugee Women
tell their story, 2013” refers to women only and is confined to the women
surveyed in that study, “Women surveyed [in that study] had served an average
of five years”. Furthermore, there are many more sources that describe the
norm period as being lengthy and protracted. The UKFFM materials record some
examples. Viewing the evidence as a whole, we consider that the position taken
in the two UNCOI Reports is broadly reflective of the bulk of the evidence. The
2016 Report states at [206] that national service is “routinely well beyond the
18 months provided for in the 1995 decree, and frequently for periods exceeding
well over a decade”. Although we have not accepted the view expressed in both
UNCOI Reports that release from military/national service is rare, we cannot
ignore the very considerable body of evidence indicating that the duration of
national service is protracted. We find telling the fact that (as noted in the
AI “Just Deserters” Report in Part 1) the Wall Street Journal, whose
correspondent was permitted a media trip to
Eritrea
in September 2015, reported
that the Eritrean government had rejected a $222.7 million plan from the EU to
facilitate the demobilisation of long serving conscripts because “it would
violate the principle that no one is exempt from patriotic duties”.
306.
We
are bound to say we have had very considerable difficulty deciding this issue,
notwithstanding the preponderance of sources that describe national service as
protracted, for two reasons. First, because for reasons set out earlier we
consider it likely that release is commonplace. Secondly because (as also noted
earlier) the figures of persons involved in national service at any one time appear
to indicate that 9 out of 10 persons are not engaged in national service
duties. If we had felt able to draw inferences from these two findings alone,
we might well have concluded that the Eritrean authorities are likely to regard
7 years as being long enough for them to be satisfied an Eritrean citizen has
completed national service. We are certainly satisfied that the great majority
of Eritreans begin national service at the age of 18 (if not earlier) and
continue in national service beyond the 18 months period and that this means
that ordinarily, by the time they reached 25 (if they have not been discharged,
dismissed or released), they would have performed 7 years of national service.
As a corollary, we would have concluded that the category of those who have
left Eritrea
illegally who would be perceived on return as draft evaders or
deserters would be confined to those who were under the age of 25 or could
otherwise show that they had not yet served 7 years. However, we do not think
inferences can be drawn from these two findings alone. It seems to us that the
broader body of evidence identifying national service as prolonged must be weighed
in the balance and accorded due weight. Even in relation to the evidence
regarding release, it is likely that in a significant number of cases release
is simply de facto, without it being confirmed by official documentation
which makes it likely that it would be difficult for the generality of beneficiaries
to show that their national service was formally complete.
307. We find it very striking that not more attention has been paid to the fact that 9 out of 10 persons are not engaged in national service duties by country analysts. We do not exclude that further information may become available in the future making clearer what the position is, as regards completion of national service, for such persons. It may be that this could vindicate our hypothesis that the average period for completion of national service is 7 years. But on the basis of the evidence before us, this seems to us a classic example of a situation where we should not depart from existing country guidance as set out in MO on this matter for the reason articulated by the UT in EM & Others [2011] UKUT 98 at [72] that “any assessment that the material circumstances have changed would need to demonstrate that such changes are well established evidentially and durable.” In short, we do not find that such a change is well established evidentially and durable.
Eligibility for national service and exit visas
308.
By
Article 17 of Proclamation No.82/1995 an Eritrean citizen “under the obligation
of national service... may be allowed to travel abroad” by producing evidence that
he or she is exempted or has completed his or her service or by producing a
registration card and entering into a security bond. Lawful exit from Eritrea
requires an exit visa issued by the Department of Immigration. According to the
2015 UNCOI Report at [401]-[413] exit visas are issued to certain individuals
without difficulty and in this regard mention is made of three categories:
older women; individuals who have completed national service when the nature of
their occupation requires regular travel; and conscripts travelling for
official business for the government, although it is emphasised that the system
operates arbitrarily.
309. In MO at [106] the Tribunal endorsed PK’s identification in 2011 of seven categories of lawful exit:
(i) a male of 54 years or over;
(ii) a female of 47 years or over;
(iii) children of seven or younger;
(iv) a person declared by an official committee to be unfit on medical grounds to perform any military or national service;
(v)
a
person certificated by an official committee to be unable to receive
appropriate medical treatment in Eritrea
;
(vi) highly trusted government officials and their families;
(vii) members of ministerial staff recommended by the department to attend studies abroad.
310. The EASO Report states (at [6.4]) that “most sources agree that exit visas are generally issued to the following categories of persons. However, some contradictions and uncertainties remain, particularly regarding the age boundaries”. Its list is as follows:
· Men aged over 54
· Women aged over 47
· Children aged under 13 (some sources state an even lower age)
· People exempt from national service on medical grounds
· People travelling abroad for medical treatment and in individual cases for studies or for a conference
· In some cases, businessmen and sportsmen
· Former freedom fighters (Tegadelti) and their family members
· Authority representatives in leading positions and their family members
Submissions
311.
The
appellants submit firstly that the position in relation to exit visas for those
within the national service scheme has not improved since MO and that
the exceptions therefore remain limited. Secondly, they submit that there is
credible evidence that there has in fact been a narrowing in the age range of
those able to obtain exit visas, “such that with the limited exceptions still
applicable, those under the age of 5 and over the age of 70 regardless of their
gender, are unlikely to leave Eritrea
lawfully”. The appellants also dispute
the respondent’s claim that around 60,000 - 80,000 persons are granted exit
visas. Mr Knafler said that the only source for this was a statement by
Eritrean immigration officers to the UKFFM and this should be given little
weight. In any event, submit the appellants, this figure was not an indicator
of Eritreans seeking to leave
Eritrea
to go abroad, since it must include
substantial numbers of Eritreans who go back to
Eritrea
for holidays.
312. In submissions UNHCR cites the list of categories as set out in the EASO Report and comments that the ability of some of these categories of individuals to obtain exit visas was recently confirmed in the UNCOI Report. UNHCR mentions ‘older women’, ‘individuals who have completed national service when the nature of their occupation requires regular travel’, and people travelling for medical reasons. The UNHCR Evidence Table notes that according to Landinfo and the EASO Report, exit visas are generally only available to women aged over 47 and according to the 2015 UNCOI Report there appears to be a “general travel ban enforced on children” (at [411]).
313.
The
respondent’s position is that lawful exit remains a real possibility and it
cannot be assumed that an Eritrean applicant for protection left illegally. According
to the respondent, the figure given by an Eritrean official to the UKFFM of
60,000 - 80,000 visas per year is not inherently fanciful or unrealistic. Far
from it having become more difficult since MO to obtain an exit visa (as
contended for by PK among others), there is evidence pointing in the opposite
direction, to some relaxation. The immediate reason why the Tribunal in MO
considered the categories were limited to seven was a temporary suspension of
exit visas and passport services in August 2008 with only partial re-opening at
the time of hearing; that no longer applied. The USSD Reports covering 2012
and 2013 had reported some relaxation, “including for medical purposes,
allowing an unknown number of persons below the age cutoffs to leave the
country” (‘Eritrea
2013 Human Rights Report’ USSD,
2014 at page 13).
314. The respondent considers relevant the data concerning European visa applications by Eritreans which although charting a drop in 2014 and 2015, showed an increase from 2011 (when they were 1789) to 2043 in 2015. The evidence justifies, according to the respondent, a reformulation of the categories able to obtain exit visas. This is similar to the EASO list except that the upper age for women should now be reduced from 47 to 30 and those granted an exit visa to travel abroad for treatment, conference and studies should be listed individually. The respondent does not accept that the evidence regarding the people’s militia justifies extending the upper age limits for men and women eligible for an exit visa. She asks us to rely on the fact that the USSD for 2015 continued to give an age level for men of 54 years and indeed for women has lowered it to 30. As regards the age limit for children, the respondent acknowledges that the same USSD Report states that “[a]uthorities generally did not give exit visas to children aged 5 or older” but notes that all recent USSD Reports including this one refer to ‘adolescents’ being granted exit permits. The respondent refers to Landinfo in April 2015 citing a source “who considered that children up to 13-14 years can receive travel documents in family reunification cases”. In the respondent’s view, this evidence justifies a higher upper limit and at the very least should persuade the Tribunal to leave the position unchanged.
315. The respondent takes particular issue with PK’s evidence that it has become more difficult to obtain a ‘medical visa’, evidence which appears to be purely anecdotal and is difficult to square in any event with his own evidence about many Eritreans travelling to Sudan for medical treatment. As regards students, the respondent also considers significant the evidence of Dr Tanja Müller recorded in 2012 that “students are being sent abroad again on scholarships for masters or PhD degrees, a programme that had all but stopped in the last decade”. Even if the Eritrean government was restrictive in granting exit visas for students to western countries, there was recent evidence of students going to countries such as China, Japan, South Korea, Dubai and the Gulf States, South Africa, India and African countries such as Kenya and Sudan. The respondent produced a number of media items referring to Eritrean students in non-Western countries.
Our assessment
316.
We
pause to remind ourselves of the unusual nature of the task we face in deciding
this issue. Ordinarily a state’s exit visa categories would be officially
declared and known. Not here, and we are faced with the unsatisfactory position
in examining the issue of having to try and construct what is likely to be the
actual list applied in practice in Eritrea
, taking into account, inter alia,
what Eritrean government officials have said about them, bearing in mind of
course that government representatives may not be stating facts and officialdom
in
Eritrea
can act arbitrarily. With that caveat, we consider that the list of
categories of lawful exit given in the EASO Report provides a more useful
starting point than that constructed over five years ago by the Tribunal in MO,
when one of the bases for it was the lack at that time of a full lifting of the
2008 suspension of exit visa and passport service.
317. We do not accept that the introduction of the people’s militia in 2012 has raised the age limits for exit visas beyond the national service age limits of 54 for men and 47 for women. Whilst there is significant evidence going both ways on this issue, we are not persuaded that it can be concluded that the authorities treat eligibility for the people’s militia as a barrier to obtaining exit visas. We accept that the 2015 UNCOI Report at [87] states that most Eritreans discharged from and into the people’s militia are not able to obtain an exit visa. Yet the very detailed study of the people’s militia by Lifos (Subject Report: People’s Army, 23 November 2015), having noted that women are involved in the people’s militia to a lesser extent and the conditions under which they are excluded are unclear, states that “[i]t should be noted that, there seem to be other circumstances which allow a person to be released from service”. We recognise as well that in the UKFFM interviews government minister Yemane Gebreab appears to assert at one point that people serving in the people’s militia cannot leave the country. However, (leaving aside that we treat Eritrean government sources with great caution) his evidence may be referring only to persons on active people’s militia duty (his answer to the question whether such persons were allowed to leave the country was “No, not now”). Despite the people’s militia having been in existence since 2012, the USSD Reports and the EASO Report have not seen its introduction to alter these age limits.
318. What seems particularly important to us is the fact that the people’s militia is a part-time obligation and is not established in all regions of the country (the EASO Report at 3.9 states that the People’s Army takes place primarily in Asmara).
319. Unlike the national service system it is a system that has grown up outside any legislative framework. Whilst we are prepared to accept that persons actively involved in doing people’s militia service or known to face immediate call up to the people’s militia in order to do weapons training or guard duty etc. may face refusal, we do not consider that outside of this context the authorities see time-limited trips abroad as at odds with the orderly functioning of the people’s militia.
320. We found PK’s oral evidence about the people’s militia unimpressive. It was anecdotal and his source evidence was lacking. He said that the fact that someone was not in the people’s militia did not mean that they were exempt and “it would only be a matter of time before they were holding a gun.” Despite PK’s evidence before us that the upper age limit for exit visas should be increased, he had not stated this in his February report, and his explanation for this, that he “probably made the changes as a result of the research [he] did”, is unsatisfactory considering the people’s militia was introduced in 2012 and the source he cited was a book published in 2013. His evidence about the consequences of fleeing from the people’s militia was that they are likely to be the same, but he accepted this was speculative. He also stated that his aunt was in the people’s militia. He had not mentioned her in his reports, but she is aged 40 and in the people’s militia, she was not engaged on a full-time basis but had to be available whenever there was a specific task and can be called anytime. This evidence is of a piece with other evidence establishing that the people’s militia has limited reach, it is arbitrary and episodic. It follows that we conclude that there is no reason to consider that the upper age limit for exit visas has increased.
321.
Equally,
however, we are not persuaded by the respondent’s submission (also reflected in
the policy position set out in the Home Office Country Information and
Guidance: Eritrea
: Illegal Exit, Version 3.0, August 2016) that we should
treat the age for women as having reduced from 47 to 30. We acknowledge that
30 is the age given in the USSD Report covering 2015, but the weight of the evidence
continues to indicate that the age limit is 47 and that even though a
substantial number of women are able to obtain a de facto discharge from
military service based on pregnancy, marriage, and responsibilities for
children, this does not appear to translate into them receiving the necessary
official documentation needed for exit, although clearly some, perhaps a not
insignificant number, are able to obtain this through family connections and/or
bribes.
322.
As
regards children, we do not consider that the EASO formulation at [6.4] –
“[c]hildren aged under 13 (some sources state an even lower age” (or the Home
Office Country Information and Guidance: Eritrea
: Illegal Exit, Version
3.0, August 2016 position to similar effect) - reflects the significant number
of sources who place the age at 5. As with so many aspects of the Eritrean
state, there is no certainty over whether the Eritrean government uses a
precise age or what it is - although the immigration officials told the UKFFM
it was 5. However, the latest USSD Report puts the age at 5 and we think that
best reflects the weight of the present evidence indicating that the Eritrean
government believes it has lost too many of its youth to emigration/flight: if
that is so, it is likely that it would view exit visa applications from
children with greater scepticism. The respondent correctly observes that this
same report also refers to “adolescents” being granted exit visas, but it
appears to us most likely that this phenomenon is closely linked to the context
of applications for family reunification abroad (as suggested by the Landinfo
evidence) rather than a general raising of the minimum age.
323. As regards EASO’s two medical categories (people exempt from national service on medical grounds and people travelling abroad for medical treatment), we see no good reasons to consider they have been narrowed in scope or withdrawn. Whilst PK’s evidence was strongly to the effect that they have narrowed, it was not substantiated. We find it significant (at least in respect of the latter category) that he also referred to the very sizeable numbers of Eritreans travelling to Sudan for medical treatment. Whilst he appeared to describe them being able to do so on travel permits rather than exit visas, we do not think the Eritrean authorities would adopt a stricter approach to one rather than the other. His evidence was that crossing the border into Sudan for medical treatment was relatively commonplace, but that you would only be granted a travel permit if outside draft age (applying the upper and lower limit as he put forward in his evidence). He talked of buses taking people across the border for this purpose and when he was asked whether the occupants of the buses were all over the age of 70/60 and under five, he then stated that the border was porous and in places there were no check points. PK’s evidence generally about the narrowing of the regime’s approach to exit visas was anecdotal and largely uncorroborated.
324. Similarly with students, and again notwithstanding PK’s evidence, we see no sound basis for regarding this category as having narrowed. PK was unable to justify his assertion that the number of scholarships has “diminished considerably” (PK’s Report of 4 April 2016). We consider that what was noted in MO - that it is highly likely that the Eritrean authorities have more confidence that students they allow to go to non-western countries will not defect or fail to return – remains true, particularly given the evidence we have of there being a diverse number of different non-Western countries (including Sudan) where Eritrean students go for studies. We also concur with the respondent that it makes sense to list this as an additional category on its own.
325. Nor are we persuaded by PK’s evidence that there has been any narrowing of the other categories. The EASO category of businessmen and sportsmen is corroborated by the UNCOI observation that Proclamation No 82/1995 allows for conscripts to show a registration card and leave a bond to obtain a visa and it has been made available only to conscripts travelling for official business for the government (see [407] of the UNCOI 2015 Report). It seems to us that the final category (“Authority representatives in leading positions and their family members”) likewise remains a sufficiently accurate reflection of the overall state of the evidence. In particular, we find PK’s suggestion that this should be narrowed to those in the President’s inner circle to be too dependent on anecdotal evidence.
326. Of course, in regard to all these categories we accept there are continuing uncertainties and contradictions (as highlighted by the EASO Report) and a certain degree of arbitrariness (as highlighted by PK and the UNCOI Reports). These categories represent therefore only those mostly likely to be available; there remains the possibility in any individual case of denial.
327.
We
also think that these categories are being exercised by a significant number of
Eritreans. We agree with the respondent that in approaching this issue sight must
not be lost of the fact that even on the highest figures given, a very large
proportion of the Eritrean population are not involved in national service. The
available data concerning European visa applications by Eritreans seems to us a
significant indicator of demand and we continue to think that Eritreans are
unlikely to go to the trouble and expense of applying for visas from other
countries if they do not have a reasonable expectation of being able to obtain
an exit visa (the latest version of the Home Office CIG on Illegal Exit at
2.2.3 cites the reference in the US State Department Report published in June
2015 to “the prohibitive cost of passports deter[ring] many
citizens from foreign travel. It costs a citizen in national service the
equivalent of 40 percent of his or her gross yearly salary to obtain a valid
passport. Some persons previously issued passports were not allowed to renew
them, nor were they granted exit visas”. At 7.1.1 is also noted that exit visas
cost 200 nakfa and are valid for one month and one trip out of the country).
This data shows a drop in 2014 and 2015 but even so an increase from 2011 (when
they were 1789) to 2043 in 2015. Such evidence certainly does not suggest a
narrowing of exit visa categories. We also consider salient the
evidence of the diverse number of non-Western countries which have Eritrean
students, in our view another strong indication that a significant number of
Eritreans regard the categories of lawful exit as offering real possibilities
for them even for those of draft age par excellence. We
do not however, place reliance on the figure given for the UKFFM by an Eritrean
immigration officer of 60,000 - 80,000 applications per year for exit visas as
it is not corroborated by any independent source. (We would note that we are
not persuaded that the exit visa figures would include persons who go to Eritrea
on holiday. Not only would including them make no sense of the figures, since their
numbers appear to add up to more than the total figures otherwise, but it also seems
to us odd that, if an Eritrean abroad makes contact with an Eritrean Embassy,
pays the 2 per cent tax and (if relevant)
signs the regret letter and gets an ID document, he or she would not receive a
stamp on arrival that would automatically allow them to exit within a certain
period without further need for an exit visa.)
328. We conclude that the categories of lawful exit have not significantly changed since MO. The Eritrean system of exit visas continues to afford, and to be perceived by a significant number of Eritreans as affording, real, albeit restricted, possibilities for them to avail themselves of and accordingly we would list the exit categories as follows (where the categories are different from those given by EASO, they are underlined):
· Men aged over 54
· Women aged over 47
· Children aged under five (with some scope for adolescents in family reunification cases)
· People exempt from national service on medical grounds
· People travelling abroad for medical treatment (this is now listed as a separate category)
· People travelling abroad for studies or for a conference [This is now listed as a separate category. We do not think that the EASO qualifier “and in individual cases” serves any descriptive purpose]
· Business and sportsmen [here again we do not think that EASO’s prefatory words “[I]n some cases” adds any descriptive purpose]
· Former freedom fighters (Tegadelti) and their family members
· Authority representatives in leading positions and their family
The 2 per cent tax and the regret letter
Submissions
329.
The
respondent’s position, based inter alia, on the DFFM Report (including
the Eritrean MoFA source who had stated that national service evaders and
deserters have the possibility of restoring their relations with the Eritrean government
by paying this tax and signing the letter of regret, the Landinfo Reports and
the Home Office CIGs), is that Eritreans can return safely to Eritrea
should
they pay the tax and sign the letter of regret (also known as the letter of repentance
or apology or rehabilitation) notwithstanding that they left illegally and/or
evaded or deserted from national service. The appellants’ position and that of
UNHCR is that this is entirely misconceived and that there is no evidence to
support the respondent’s case and that the tax and letter are used by the
diaspora in order to access consular services only.
330.
We
were referred to findings of the 2015 UNCOI Report at [440]) that many
Eritreans no longer have an Eritrean passport and can only obtain one after the
payment of the 2 per cent tax collected through Eritrea
’s diplomatic
representatives abroad. In order to ensure the payment of the tax, methods used
have been found to be illicit by the United Nations Security Council. The
Commission found that one of the methods used is to force payment for basic
consular services and that the non-payment of the tax presents a security risk
for arrest and detention (see [441]). In addition to the tax, according to the
Commission in the same report at [442], Eritreans who left the country illegally
have to sign a document to regularise their status before they can request
consular services. By signing the form an individual admits the offence and
agrees to accept punishment.
331.
We
were referred to the anonymous source 2 spoken to by the UKFFM. This source did
not assert that the payment of the tax would enable a person to return safely,
but rather linked it with renewing licences for family members in Eritrea
. Anonymous source 1 said that those who left illegally pay 2 per cent tax “and
come back legally” going on to say that they visit their family here and then
return to the country from where they came.
332. We were referred by Mr Rawat to the evidence from the 2015 UNCOI Report (see [436]) concluding that there are two exceptions to the rule that returnees are arrested, detained and forced to enlist in national service following a forced return. The report refers to a group of Eritreans who returned from country “D” with a letter certifying that they had paid a 2 per cent rehabilitation tax having already been detained for several years in country D. Specific reference is made to an individual from the group who had been imprisoned for three years in country D. Another case is referred to of forced repatriations in 2014 where seven older men were reportedly freed whilst younger men who were also forcibly returned at the same time were not released.
Our Assessment
333. The weight of the evidence points very much in the direction that the letter and the tax do not guarantee safety for Eritreans returning; rather they enable them to access consular services. There is scant evidence of anyone who has not been naturalised in another country paying the tax and/or signing the letter and returning safely or otherwise. We accept PK’s evidence about this, which was very much corroborated by evidence from other sources. There being insufficient detail about the returnees to draw conclusions, we would have reached this conclusion independently in any event. Apart from the two exceptions referred to by the UNCOI, it would appear that the bulk of the examples cited concern or may concern persons who voluntarily returned, which in our view (as set out below when dealing with failed asylum seekers and forcible returns at [335]-[337] and [357]-[367]) puts them in a different category.
334. Suffice to say for the purpose of this section, that we do not accept that the evidence goes anywhere close to establishing that the payment of the tax and the signing of the letter would enable draft evaders and deserters to reconcile with the Eritrean authorities. In relation to the letter of regret, we also have serious doubts that it can properly be described as a basis for reconciliation, since its terms amount to a confession of guilt by the person who signs it to what the Eritrean regime considers “appropriate punishment” in the context of a regime with a very poor human rights record.
Failed Asylum Seekers
335. In MO the Tribunal at [131] held that failed asylum seekers as such are not at risk of persecution on return. We do not detect any enthusiasm from any of the parties for a different view being taken today. Indeed the appellants’ expert witness PK, was adamant that failed asylum seeking could not be enough on its own to engender risk because of the main reasons highlighted in MA and MO that the Eritrean authorities have a vested interest in embedding abroad people who claim asylum but are in reality well-regarded by the government and that a significant number appear to be in reality supporters of the Eritrean government or able to demonstrate that they are through attendance at rallies etc.
336. We note that references can be found in some of the sources taking a different view, but here we regard the way the matter was put by the April 2015 Landinfo Report, that there was ”no empirical evidence” to support the contention that an application for asylum will lead to adverse reactions from the Eritrean authorities, as being entirely fair.
337.
To
the extent that any inferences can be drawn from the evidence overall, it seems
to us that there is likely to be a further reason presently why the Eritrean
authorities would not view the mere fact of being a failed asylum seeker
adversely. This is that the Eritrean authorities consciously recognise the
economic value to them of having a sizeable diaspora who send remittances and
some of whose members also pay the 2 per cent tax. Rightly or wrongly, they
clearly consider that many of the Eritreans who have left have done so out of a
desire for economic betterment rather than asylum yet go on to claim asylum as
a way of residing elsewhere. That may be a factor that has played a part in
Eritrean government thinking for some time, but recent evidence does underscore
how greatly the Eritrean government depends on foreign remittances. According
to Crisis Group Africa Briefing No 100 August 2014 (“Eritrea
: Ending the
Exodus?”) remittances inject hard currency into the country’s meagre foreign
exchange reserves, whilst bolstering the economic resilience of the families
left behind and the government has become increasingly dependent on Eritrean’s
abroad as a source of capital. It was estimated that approximately one third of
Eritrea
’s 2005 GDP came from remittances and this may have increased. Whilst
there are still references in some sources to the Eritrean authorities viewing
failed asylum seekers as traitors, we continue to follow MO in
considering this as something only likely to be acted on in any way when there
is a particular symbolic importance for
Eritrea
public policy e.g. when dealing
with collective expulsions back to
Eritrea
. This last observation, however, is
we think of greater importance than previously, because what we have to
consider is not just how failed asylum seeking as such would be perceived, but
how the Eritrean authorities would react to persons perceived as draft evaders
or deserters when forcibly returned.
Illegal exit by those perceived on return to be draft-evaders or deserters
Submissions
338. The respondent maintains that according to the DFFM and UKFFM interlocutors the Eritrean government does not detain or punish evaders and deserters within the country systematically and was more concerned to put them back in national service work. She observed that the 2015 UNCOI Report at [818] noted the grant of an amnesty to deserters in November 2014. The evidence regarding the “shoot to kill” policy indicates it was less in evidence and that round-ups or giffas are less frequent. The evidence shows, she submits, that targeting of relatives had also reduced.
339.
The
respondent also maintains that in any event it is incorrect to infer that those
who have left the country would be punished in the same way as those caught
within the country. That is because the latter can obtain an Eritrean passport
and/or ID card by paying fees, the diaspora tax and signing a letter of
apology. Reliance was placed on the Eritrean MoFA who had stated to the DFFM
that national service evaders and deserters have the possibility of restoring
their relations with the Eritrean government by paying this tax and signing the
letter of apology. It was indeed this view that led the respondent in its
March 2015 CIG Policy Eritrea
: Illegal Exit the Home Office, drawing
heavily on the DFFM Report, to decide that the guidance given in MO
should no longer be followed, stating at 1.3.4 that:
“However MO was promulgated
in 2011. The most up-to-date information available from inside Eritrea
notably
the [DFFM] Report.....[indicates that as] a result Eritreans who left illegally
are no longer considered per se to be at risk of harm or mistreatment amounting
to persecution on return. Consequently, the guidance outlined in MO
above should no longer be followed and failure of a person to comply with a
reasonable request to pay diaspora tax would not in itself give rise to a well-founded
fear of persecution or serious harm.”
340. The respondent also submits that illegal exit is no longer a risk factor or as great a one as before - because there is now a body of evidence showing that each summer many members of the Eritrean diaspora return for holidays. Western embassy A gave an account to the DFFM of 400 Eritreans with Swedish passports being stranded in Asmara following the collapse of an air company. The majority had left illegally yet were all able to return to Sweden. According to the NCEW source, there were also diaspora returnees who came back to settle and set up businesses. The respondent considered that the cohort of diaspora returnees must include those who left illegally, yet there was no evidence of such individuals being subjected to ill-treatment.
341. The respondent also seeks to rely on the fact that the Eritrean immigration officials interviewed by the UKFFM said that those who had been outside the country for three years or more were free to return. If they came back within the three years they had to complete national service. Those interviewed by the UKFFM included persons who had not been naturalised in their destination countries. For the respondent the evidence from the UKFFM necessitated, in consequence, a more fact-specific analysis than presently required under existing country guidance.
342.
The
appellants’ counter-argument was as follows. Firstly, it was said that the
evidence of what happens in-country to draft-evaders and deserters was still
overwhelmingly to the effect that they were the recipients of ill-treatment.
Secondly it was argued that statements made to the UKFFM by officials and
others that there is some kind of amnesty for deserters/evaders who pay the tax
and/or sign the letter of regret was “blatantly untrue” considering that [84]
of the 2016 UNCOI Report showed that the Commission had received reliable
information indicating that the office of the President had instructed Eritrean
officials meeting delegates to make certain assertions. Thirdly, the appellants
submitted that it was highly unlikely that those who had fled Eritrea
and are concerned about punishment directed towards them or their family members on account
of illegal exit would pay the diaspora tax and make their whereabouts known to
the authorities. PK’s evidence was clear that the diaspora tax was not
something that was in fact paid by persons likely to face forcible return and
in any event, paying it would not immunise them from ill-treatment on return
because that tax gave one access to domestic services but did not extinguish
the fact that they would be perceived on return as evaders and deserters and
punished accordingly. Fourthly, even if the diaspora tax can somehow immunise
those who exited illegally from persecution on return, these are sums extracted
from members of the diaspora, often by illegitimate pressure to fund purchases
of arms in breach of a UN arms embargo and militant groups that destabilise the
region. Further, it could not be suggested consistently with HJ (Iran) [2010] UKSC 31 and RT (Zimbabwe) [2012] UKSC 38 that a person who is unwilling
to subject themselves to national service or the people’s militia or otherwise
unattractive aspects of the Eritrean regime, should not be treated as in need
of protection simply because they could take those steps. Fifthly, as regards
signing of the letter of regret, the appellants’ submission was that it does
nothing more than express consent to whatever punishment the government
considers fit for desertion or draft evasion. “It is very unlikely that an
individual who has been subjected to past persecution would consent to such
treatment and in relation to forced returnees, it seems entirely irrational
that an individual would sign”.
343. UNHCR’s submission reminded the Tribunal that in the UNHCR Eligibility Guidelines of 2011, which UNHCR say continue to apply, it is stated that “[d]raft evaders/deserters are reported to be frequently subjected to torture”. Desertion and draft evasion were criminal offences under Eritrean law. The UNHCR submissions placed reliance, inter alia, on the statement by the USSD of June 2015 that “refusal to perform military or militia service, failure to enlist, fraudulent evasion of military service and desertion were punished by lengthy imprisonment or other arbitrary forms of punishment” and similar observations made by the UN Special Rapporteur, the UNCOI and by AI and HRW, the latter who wrote in July 2015 that:
“The preponderance of evidence...
indicates that there has been no change in Eritrea
’s treatment of draft
evaders, deserters and people leaving the country without permission.”
Our Assessment
344. As regards the issue of how decision-makers should decide whether a person has left illegally, we see no reason to differ from the precise terms of the guidance in MO at (iii):
“(iii)....The general position as
regards illegal exit remains as expressed in MA, namely that illegal
exit by a person of or approaching draft age and not medically unfit cannot be
assumed if they have been found to be wholly incredible. However, if such a
person is found to have left Eritrea
on or after August/September 2008, it may
be that inferences can be drawn from their health history or level of education
or their skills profile as to whether legal exit on their part was feasible, provided
that such inference can be drawn in the light of adverse credibility findings.”
None
of the parties has pointed to any evidence indicating the need for a different
approach on this issue. We would next reiterate that it is incorrect of the
March and September 2015 and August 2016 CIGs to portray (as they certainly do in
places) the position set out in MO as being that Eritreans who left
illegally are considered to be, per se, at risk. The MO position
was explicitly stated as being subject to three exceptions (see [133iv]).
Indeed, UK country guidance has never asserted that the fact of illegal exit
from Eritrea
is of itself enough to place a person at risk.
345. We are bound to say that certain of the arguments advanced seemed to us to obfuscate rather than assist the Tribunal’s task. The possibility canvassed by the appellants, for example, that sums extracted from members of the diaspora may be used by the Eritrean government to fund purchases of arms in breach of a UN arms embargo by militant groups that destabilise the region, seems to us far removed from the task of identifying risk categories or factors. Be that as it may, our view is that the totality of the evidence continues to support the view that the fact of illegal exit is not of itself enough to place an individual at risk.
346. The question is, therefore, what further characteristics are needed to place a person at real risk of persecution or serious harm on return.
347.
We
consider two further characteristics are needed: (i) that they will be perceived
on return as evaders/deserters; and (ii) that they will be persons subject to
forcible return. Even then, however, we continue to think that this category is
subject to certain exceptions and that they are exactly the same as those
identified in MO, namely (1) persons whom the regime’s military and
political leadership perceives as having given them valuable service (either in
Eritrea
or abroad); (2) persons who are trusted family members of, or are
themselves part of, the regime’s military or political leadership. A further
possible exception, requiring a more case specific analysis is (3) persons (and
their children born afterwards) who fled (what later became the territory of)
Eritrea
during the War of Independence. We do not accept the position identified in the latest
version of the Home Office CIG on Illegal Exit published on 4 August 2016 that
the scope of these exceptions has widened.
348.
The
respondent has sought to argue that we should adopt a more open-ended
fact-specific analysis, but her argument is dependent on the premise that those
who have left Eritrea
illegally as evaders or deserters have the ability to
regularise their position by payment of the diaspora tax and letter of regret.
For reasons given above at [333] we reject this. Persons who are likely to be
perceived as deserters/evaders will not be able to avoid exposure to such real
risk merely by showing they have paid (or are willing to pay) the diaspora tax
and/have signed (or are willing to sign) the letter of regret.
349. Whilst we accept there are individual examples of the government not punishing or mistreating returning draft evaders or deserters who left illegally, they are small in number and in some cases appear to pertain to those who returned voluntarily (and may have naturalised in another country: see immediately below).
350.
Insofar
as the evidence of diaspora members returning to Eritrea
for holidays is
concerned, the evidence does appear in one respect to support the respondent’s
position in that it persuades us that they include a significant number of
draft evaders and deserters who left illegally. We infer that because if it
were confined to those who left
Eritrea
during the war of independence and
their children and/or those who left lawfully, that would have been identified
by more than one source on the basis of some empirical evidence. Given that the
annual numbers appear relatively high (even if not as high as the government
figure of 80,000-90,000 as stated by Yermane Gebreab), we seriously doubt that
all could be from that category. Even according to PK (whose evidence was the
main source for the view that they were confined to this cohort) there are
significant numbers of individuals among the diaspora who left
Eritrea
illegally after 1991 but who have close connections with the present government and
as such would be unlikely to face any difficulties on return.
351. However, it seems to us that the great majority of such persons are likely to be naturalised. We accept that the evidence regarding this is sketchy, but consider it a reasonable inference that (unless having close connections with the present government as set out in the first two of three aforementioned exceptions) persons who have not naturalised would not put themselves and their families in the position of going back to a country with such a poor human rights record. Such an inference would clearly not be warranted if there was concrete evidence of persons (evaders/deserters) who exited illegally going back without having naturalised; but there is not.
352.
One
of the main sources relied on by the respondent regarding this matter, the DFFM
Report, is based to a significant degree on evidence given by PK, which he says
was misinterpreted and in consequence cannot be relied on. The same we think
is true of the evidence of Eritrean government interlocutors consulted by the
DFFM team. It must also be borne in mind when considering the possible identity
of those who go back for holidays, that diaspora persons who have not
naturalised will certainly include those who have obtained settlement or some
kind of permission to stay under the Immigration Rules, but they will also
include asylum seekers or illegal entrants or overstayers. It is difficult to
see that any persons in the last three categories would voluntarily leave the UK to go anywhere, let alone Eritrea
, since they would thereby negate their chance of returning. It is
also a reasonable inference that a significant number of those who have
acquired refugee status (but who have not naturalised), would be wary of returning
to
Eritrea
on holiday for fear that such conduct may expose them to cessation
action on return back to the country of refuge (although we cannot assume, we
accept, that all of those we are discussing necessarily act according to such
concerns).
353. The significance of holidaying returnees having prior naturalisation is that, whilst the Eritrean government might well have chosen to disregard their foreign nationality and rely simply on their being Eritreans who left illegally and who are draft evaders/deserters, it appears very much that they do not, as the Western embassy example given to the DFFM illustrates.
354.
What,
then, is the basis for considering that those who left illegally and will be
perceived on return as draft evaders/deserters would be at risk? There is first
of all, the evidence as to what happens to evaders/deserters within Eritrea
. As explained at [253]-[256], we are satisfied that despite a lessening in the
frequency of round-ups (giffas) and “shoot to kill” operations and punishment
of relatives, the treatment such persons are likely to face amounts to
persecution or serious harm, since it continues to take the form of widespread
recourse to detention. Mr Rawat conceded during the hearing that anything more
than a very short period in detention in
Eritrea
would carry a real risk of
ill-treatment and on the available evidence there is in our judgement a real
risk that draft evaders/deserters regularly face more than very short-term
detention. There is some evidence that some persons may, instead of detention,
face assignment to military/national service, but for an initial period of
time, it is likely this will be assignment to military duties and, in any
event, as will be explained below, we consider that a requirement to perform
national service duties, military or civilian, would constitute forced labour
contrary to Article 4 of the ECHR, if not also Article 3.
355.
Second,
argument that the Eritrean authorities would treat returning evaders/ deserters
differently from in-country evaders/ deserters seems to us insufficiently made
out. Indeed, one of the most recent sources cited at 11.1.26 of the August CIG
on Illegal Exit (the Swiss Report of March 2016) states that “[t]he few
available reports indicate that the authorities treat them similarly as persons
apprehended within Eritrea
.” This brings us to the second characteristic which
we consider is required to bring a person within a risk category.
356. The specific category of persons with whom we are concerned are not draft evaders or deserters who have left illegally and would be making a voluntary return. In relation to the latter there are some possible examples in the evidence which suggest they can reach reconciliation with the Eritrean authorities. We have taken particular note in this regard of the sources relating to voluntary returns cited by the latest version of the Home Office CIG on Illegal Exit at 10.1.16 (citing a UKFFM source), 11.12.1 and 11.1.26 (citing the Swiss Report on March 2016). Those with whom we are concerned are persons who are or will be perceived as evaders/ deserters and who will be known to be persons who are the subject of a forcible return. Whilst we do not necessarily think the Eritrean authorities would react in precisely the same way to individual forced returnees as they have in the past to mass forcible returnees, we consider it reasonably likely that they would feel similarly impelled to adopt a punitive stance in a way they have not sometimes done to voluntary returnees. On the totality of the evidence we consider this is a reasonably likely state of affairs. We must analyse the issue of forcible returns in more depth in the next subsection.
Forcible Returns
357.
In
MO the Tribunal had little evidence before it regarding individual
forced returns and was cautious of attaching weight to an AI report of two
forcible returns by Germany in 2009 in light of the shortcomings in sourcing
(see [126]). It recorded ongoing concerns about the treatment of those subject
to mass forcible returns from Malta, Libya and Egypt in the 2002-2009 period,
but seemed to proceed on the assumption that individual forcible returns to Eritrea
were an ongoing reality.
Submissions
358. The respondent’s position is that since MO the evidence about likely ill-treatment to persons forcibly returned remains vague and tenuous. She notes, as did Landinfo 2013, that PK did not have concrete information. The one example he gave concerned a Mr Berhane but his information left unclear whether this man had a particular political profile. The respondent regards UNHCR’s evidence on this issue as lacking proper sourcing and the UNCOI Report instances appear to relate to forced repatriations in 2002, 2004 and 2008 and the only two examples given in these reports that are post-MO are somewhat unclear.
359.
The
appellants’ and UNHCR’s closing submissions highlighted the recent evidence to
hand about deportations from Sudan based on several reports, including Martin
Plaut’s Report of 2 June 2016 that 900 Eritreans had been picked up in Khartoum
and refouled and that 800 were deported while getting ready to go to
Libya; the Report from Kirsty Siegfried of 25 May 2016 which says that the
authorities in Sudan have launched a crackdown on Eritrean migrants and have
summarily deported c1300 Eritreans in c.23 May 2016 and that they remain
detained in Eritrea
; and the 2016 UNCOI Report at [98] stating that Sudan
deported 313 Eritreans back to
Eritrea
on 22 May 2016 and that on arrival the
returnees were arrested and detained. Witnesses were reported as saying that
those who were in national service prior to departure were held at Abeito
Prison and those who had not undergone national service were awaiting transfer
to military training call up. (There were reports from UNHCR (see 11.1.24 of
the Home Office CIG on Illegal Exit, August 2016) and HRW which appear to cover
the same May deportations.)
Our assessment
360. We consider that the further evidence before the Tribunal since MO regarding forced returns requires us to address its implications more closely. Although there are some shortcomings in the sourcing of the evidence regarding forcible returns since 2011 (e.g. inconsistency between the various sources over the precise details regarding recent returns to Sudan), we cannot ignore the fact that the evidence suggests much more strongly than was the case in MO that draft evaders and deserters forcibly returned run a real risk of suffering ill-treatment.
361.
The
2015 EASO Report notes that Eritreans were repatriated from Egypt in 2009, 2011 and there have been many instances of overland repatriations from Sudan in recent years. “No information is available on the fate of those repatriated after
their return, however”. It goes on to note, nevertheless, that some of the
respondents contacted in Eritrea
during Denmark’s and Norway’s FFMs in late
2014 and early 2015 believed that (repatriated) deserters and draft evaders
were held in prison for several weeks or months and were then reassigned to
national service. The EASO Report records what appears to be the same incident,
noting that some of the respondents contacted during Denmark’s and Norway’s
FFMs in late 2014 and 2015 considered a spell in prison of at least several
weeks was the likely outcome for those forcibly repatriated: see para 3.8.2.
362.
The
Arapiles study (“The True Human Rights Situation in Eritrea
: The New UK Home
Office Guidance as a Political Instrument for the Prevention of Migration” p19
at n182]) refers to an Eritrean refugee interviewed in April 2015 stating that
he had been deported from country X when his student visa expired in 2013 and
suffered ill-treatment.
363.
The
2015 UNCOI Report notes at [430] the forced repatriation of around 40 Eritreans
from country E in 2014, plus an unspecified number from country D in the same
year. It is said in the same context that there is a “common pattern” of
systemic ill-treatment of such persons ([431], see also [1069]). At [433] the
report notes that several hundred Eritrean refugees who managed to escape and
were forcibly returned to Eritrea
were reported to have faced detention, torture
and other forms of inhuman treatment. At [436] the Commission said it had
found “two exceptions to the rule that returnees are arrested, detained and
forced to enlist in the national service upon their arrival in
Eritrea
”. These were:
“A group of Eritreans has returned
from [country D] with a letter certifying that they had paid the 2 per cent
Rehabilitation Tax and had already been detained several years in [country D].
The witness had himself been imprisoned for three years in [country D]. He was
given a permit to return to his home town, but which had to be reviewed every
two months. He left Eritrea
again shortly after being deported. The other
case concerned forced repatriation to
Eritrea
in 2014, where seven older men
were reportedly freed, while the younger men who were returned to
Eritrea
at the same time were not released.”
364. Albeit the recent evidence is sketchy and even though it falls short of solid documentation, in terms of the number of incidents (some very recent), it amounts to significantly more than was before the Tribunal in MO and we do not think it can be disregarded.
365.
The
UT in MO considered that the context of mass forcible returns may affect
the reaction of the Eritrean authorities because such returns are likely to
have a public profile and those authorities may feel they have to send out a
tough message. By implication the UT did not necessarily consider that those
authorities would react in a similar way to forcible returns undertaken on an
individual basis. We feel it necessary to reconsider that assumption. The
overall context is highly unusual because it would appear that in Europe and
indeed other Western countries, governments have not been making any enforced
returns to Eritrea
for some time; that at least is what we understand the
position to be from Mr Rawat when we sought clarification on the matter. That
was not the Tribunal’s understanding of the position in 2011, when, in any
event, although falling short of considering that those who left illegally
would per se face ill-treatment on return, the UT assessed that the
great majority would.
366. Be that as it may, the recent evidence of forcible returns made from non-Western countries, chiefly the overland repatriations from Sudan, is really the only type of evidence we have against which to assess risk on return from Western countries such as the UK. And it constitutes evidence showing that in the last few years those who are likely to be perceived on return as draft evaders/deserters and who have been the subject of such forcible returns have met with, or are likely to have met with, ill-treatment on return. Further, recommencement of forcible returns from Europe would very likely in our judgement be seen by the Eritrean authorities as requiring them to adopt a punitive stance even in relation to persons in the aforementioned category who are returned individually. We infer that their reaction to such a re-commencement would be a matter of high importance to the regime.
367. It is possible to conjecture that the Eritrean government would feel the need, especially in the light of recent EU funding, to demonstrate a more relaxed or softer policy, such as was mooted in the DFFM Report mainly (it seems) by reference to voluntary returnees. On the other hand, the evidence points more strongly to the policy imperatives of the current Eritrean government being driven not by concerns about its image in the eyes of Europe and the West but by domestic concerns about the maintenance of control and regulation of their own population and the need to show that those perceived as draft evaders or deserters would not receive preferential treatment on return. In our judgement there is a real risk that the likely reaction would therefore be similar to that given to those forcibly repatriated from Sudan and the evidence we have about that indicates such persons are likely to face treatment contrary to basic human rights.
Draft Evaders and Deserters
368. To this point our assessment of the issue of risk on return to those who left illegally and are likely to be perceived on return as draft evaders and deserters is not markedly different from MO. We now have to consider whether it remains sufficient that such persons have exited illegally and are of or approaching eligible draft age (unless falling within one of three specified exceptions).
369. As can be gleaned from our earlier observations when considering duration of national service ([261-263]) and discharge ([270] – [278]) we have found it very difficult to resolve this issue, particularly given that the appellants’ case and that of UNHCR is not easy to square with the figures available as regards those who are performing national service duties and the fact that for reasons given earlier we have accepted the respondent’s case before us that “discharge” from national service is commonplace.
370.
However,
for the reasons we have given earlier, we conclude that the preponderance of
the evidence continues to support the MO position and that, although it
is reasonably likely that persons who have been released will have
documentation which will enable them to travel within Eritrea
, the fact that
they are reservists (a term we use here simply to identify those who have been
discharged/released) would not entitle them to an exit visa. Whilst release is
commonplace, it appears that it is often de facto and that those who
benefit would not ordinarily be given or hold official documents confirming
that they have completed national service. We consider that recall is not
common but that the Eritrean system operates to ensure that the great majority
of those of or approaching draft age are regarded as still “on the books” and
as not having completed national service. What was noted in the EASO Report regarding
civilian national service and those in ministries strikes us as very pertinent:
“[m]any employees of ministries do not know whether they are still engaged in
national service or have been dismissed”. We remind ourselves that the great
majority of sources, including the very recent UNCOI Reports, consider that the
duration of national service is prolonged. From the evidence we conclude that
a person who exits
Eritrea
illegally and is of or approaching draft age, is
likely on return to be perceived as an evader or deserter because of
non-completion of national service.
National service as slavery or servitude or forced labour
371. We explained at [14] that one of the country guidance issues to be dealt with in this case is “(iii) Whether the Eritrean system of military service gives rise to a real risk on return of exposure to treatment contrary to Article 4 ECHR.”
372. Article 4 provides as follows:
“(1) No one shall be held in slavery or servitude.
(2) No one shall be required to perform forced or compulsory labour.
(3) For the purposes of this Article “forced or compulsory labour” shall not include:
(a) any work required to be done in the ordinary course of detention imposed according to the provisions of Art.5 of the Convention or during conditional release from such detention.
(b) any service of a military character, or in the case of conscientious objectors in countries where they are recognised, service exacted instead of compulsory military service.
(c) any service exacted in case of an emergency or calamity threatening the life or well-being of the community.
(d) any work or service which forms part of normal civic obligations.”
373. By virtue of Article 15(2), this Article admits of no derogations or limitations in respect of slavery or servitude, but the prohibition on forced or compulsory labour is a derogable provision, albeit forced or compulsory labour is only permitted in circumstances set out exhaustively in Article 4(3).
374.
For
the appellants and UNHCR, the answer to the question posed as a country
guidance issue should be that there is such a risk because the Eritrean system of
military/national service is contrary to Article 4. The respondent maintains
that no such risk arises and that in the context of extraterritorial
application of Article 4 the threshold test is higher, being that of “flagrant
denial”. The same test is posited in paragraph 2.3.44 of the Country
Information and Guidance: Eritrea
: National (incl.Military) Service, Version
3.0, August 2016.
375. At one or two points in the submissions the parties framed the conflict as being between a “flagrant denial” test and a “real risk” test, but we take their essential position - and one we discern to be clearly reflected in established case law - that the principle of flagrancy is concerned with the threshold for violation, i.e. with whether (as in the case of non-derogable rights) mere violation is enough, or whether, as in the case of derogable rights, the violation must be “flagrant”: see R (Ullah) v Special Adjudicator [2004] UKHL 26 at [28],[34],[44],[47]; EM(Lebanon) v Secretary of State for the Home Department [2008] UKHL 64 at [13],[33], [55].
Article 4 and Article 3
376. We will address the threshold issue next, but would observe at the outset that it does not seem to us to be of central importance because, to the extent that UNHCR and the appellants rely on the prohibition within Article 4 on slavery and servitude and indeed on forced or compulsory labour, their arguments are at least capable of being cast in Article 3 terms, as there would seem to be considerable overlap. In a case cited in Ullah - Ould Barar v Sweden (1999) 28 EHCRR CD 1999; 28 EHRR CD 213 - the Court found the applicant’s complaint under Article 4 (as well as his complaints under Articles 2 and 3) to be inadmissible on the facts, but recognised ‘that the expulsion of a person to a country where there is an officially recognised regime of slavery might, in certain circumstances, raise an issue under Article 3 of the Convention.’ And Lord Bingham in Ullah at [41] appeared to consider that forced or compulsory labour cases could also fall within the ambit of Article 3:
“It is no doubt right that in the modern world a case alleging slavery is perhaps a little unlikely. A case asserting forced labour is less unlikely but, if it arises, would no doubt fall under article 3.”
377. On the other hand, since neither party has sought to argue that the Eritrean system of military service was per se contrary to Article 3, we shall focus primarily on Article 4 and indeed Lord Bingham in the same paragraph [41] appeared to accept that a person seeking to rely solely on Article 4 in an extraterritorial context could not be turned away.
Article 4: the legal framework
378.
The
appellants’ submissions ask the UT to find that return of those approaching or
of draft age to Eritrea
would expose them to a breach of all three of the prohibitions
enshrined in Article 4 – against slavery, servitude and forced or compulsory
labour. UNHCR invited the Upper Tribunal to find that the return of that
category would expose them to a breach of the prohibition on servitude or of
the prohibition on forced labour not falling within the Article 4(3)(b)
exception. UNHCR also emphasised that it was not its submission that all
aspects of Eritrean national service breached those thresholds regardless of
assignment or duration,” but rather that, given the arbitrariness of duration
and assignment, there was a real risk of a breach.” It is necessary, therefore,
to have regard to Strasbourg jurisprudence on each of the three prohibitions
and their interrelationship. In Case of Rantsev v Cyprus and Russia, Application no. 25965/04, the Court noted at para 276:
“In Siliadin [Application no. 73316/01, ECHR 2005-VII], considering the scope of “slavery” under Article 4, the Court referred to the classic definition of slavery contained in the 1926 Slavery Convention, which required the exercise of a genuine right of ownership and reduction of the status of the individual concerned to an “object” (Siliadin, cited above, § 122).”
379. For the Court the concept of “servitude” entails an obligation, under coercion, to provide one’s services, and is linked with the concept of “slavery” (see Seguin v France (dec) Application no 42400/98, 7 March 2000; and Siliadin, cited above, para 124). In Siliadin the Court observed that with regard to the concept of “servitude” what is prohibited is a “particularly serious form of denial of freedom” (see Van Droogenbroeck v Belgium, Commission's Report of 9 July 1980, Series B no. 44, p. 30, paras 78-80). It includes, “in addition to the obligation to perform certain services for others ... the obligation for the 'serf' to live on another person's property and the impossibility of altering his condition”. In this connection, in examining a complaint under this paragraph of Article 4, the Commission paid particular attention to the Abolition of Slavery Convention (see also Van Droogenbroeck v Belgium Application no 7906/77, Commission decision of 5 July 1979, DR 17, p. 59).
380. The 2014 Council of Europe/European Court of Human Rights publication, “Guide on Article 4 of the Convention…” makes clear by reference to leading cases on Article 4 that “servitude is an “aggravated” form of forced or compulsory labour” and that “the fundamental distinguishing feature between servitude and forced or compulsory labour…lies in the victim’s feeling that their condition is permanent and that the situation is unlikely to change” (para 17).
381. For “forced or compulsory labour” to arise, the Court has held that there must be some physical or mental constraint, as well as some overriding of the person’s will (Van der Mussele v Belgium, 23 November 1983, § 34, Series A no. 70; Siliadin, cited above, para 117).
382. In Van Der Mussele v Belgium the Court had recourse to the ILO Forced Labour Convention (FLC) No.29 concerning forced or compulsory labour and to the fact that for the purposes of that Convention the term “forced or compulsory labour” means “all work or service which is exacted from any person under the menace of any penalty and for which the said person has not offered himself voluntarily”. The Court approved the European Commission of Human Rights’ position that:
“37. … for there to be forced or compulsory labour, for the purposes of Article 4 § 2 (art. 4-2) of the European Convention, two cumulative conditions have to be satisfied: not only must the labour be performed by the person against his or her will, but either the obligation to carry it out must be "unjust" or "oppressive" or its performance must constitute "an avoidable hardship", in other words be "needlessly distressing" or "somewhat harassing".
383. Strasbourg jurisprudence, as outlined for example, in the Van der Mussele case, understands Article 4(2) to be a right which, although derogable, is subject to very strict delimitations. It considers that Article 4(3) is not intended to "limit" the exercise of the right guaranteed by paragraph 2 (Article 4(2)), but to "delimit" the very content of this right, for it forms a whole with paragraph 2 (Article 4(2)) and indicates what "the term ‘forced or compulsory labour’ shall not include" (“ce qui n’est pas considéré comme ‘travail forcé ou obligatoire’"). This being so, paragraph 3 (Article 4(3)) serves as an aid to the interpretation of Article 4(2).
384. The four sub-paragraphs of paragraph 3 (Article 4(3)(a), 4(3)(b), 4(3)(c), 4(3)(d)), notwithstanding their diversity, are said to be “grounded on the governing ideas of the general interest, social solidarity and what is in the normal or ordinary course of affairs”.
385. In the same Guide, when considering normal civic duties and the issue of voluntary consent, it is noted that:
“…the Court will have regard to all the circumstances of the case in the light of the underlying objectives of Article 4 when deciding whether a service required to be performed falls within the prohibition of “forced or compulsory labour…”.
386. The standards developed by the Court for evaluating what could be considered normal in respect of duties incumbent on members of a particular profession take into account whether the services rendered fall outside the scope of the normal professional activities of the person concerned; whether the services are remunerated or not or whether the service includes another compensatory factor; whether the obligation is founded on a conception of social solidarity and whether the burden imposed is disproportionate (para 29).
387. Article 4(3)(c) excludes any service exacted in case of an emergency or calamity threatening the life or well-being of the community from the scope of forced or compulsory labour. This same Guide notes that in a case which concerned a requirement that the applicant serve a year in the public dental service in northern Norway, two members of the Commission held the view that the service in question was service reasonably required of the applicant in an emergency threatening the well-being of the community and was not forced or compulsory labour (I v Norway Commission decision) (para 43). In Van der Mussele the Court held, in respect of the applicant who was a pupil-advocate, that while remunerated work may also qualify as forced or compulsory labour, the lack of remuneration and of reimbursement of expenses constitutes a relevant factor when considering what is proportionate or in the normal course of business.
388. The clause excluding military service expressly prevents such service being regarded as constituting “forced or compulsory labour” per se, but there is no similar clause relating to “slavery or servitude”, hence this does not prevent it from being so regarded in some instances. As was noted by the Commission, in W,X,Y and Z application no 3435/67, “…there are historical examples of unacceptable slavery or servitude being used for purposes of military service”. However, the examination of whether any particular military service regime constitutes servitude or slavery must be informed by an understanding that the duty of a soldier to observe the terms of his service and the ensuing restriction on his freedom and personal rights does not amount to an impairment of rights which could come under the terms “slavery or servitude”.
389. There is very limited jurisprudence on Article 4(3)(c), which excludes any service exacted in case of an emergency or calamity threatening the life or well-being of the community from the scope of forced or compulsory labour, but the jurisprudence on Article 15, whose terms bear some resemblance, make abundantly clear that emergency provisions are not to be construed broadly: see P van Dijk et al (eds), Theory and Practice of the European Convention on Human Rights, 4th Ed., chapter 34.
The threshold test
390.
In
relation to the issue of the relevant threshold test to be applied in relation
to Article 4, there is a sharp divide between the respondent and UNHCR, with
the appellants’ position moving from initial agreement with the respondent to
eventual agreement with UNHCR. According to the respondent (and the appellants
in paragraph 59 of their first skeleton argument) the relevant test is whether
persons forcibly returned to Eritrea
would be exposed to a real risk of a
“flagrant breach” of Article 4 on return. According to UNHCR (and latterly the
appellants) the test is the same as under Article 3, namely whether there is a
real risk of being exposed to treatment contrary to the Article. For the test
of “flagrant breach” is confined to non-derogable rights. The respondent argues
that such an approach would be contrary to the analysis of the House of Lords
in R (Ullah) v Special Adjudicator [2004] UKHL 26, the opinions of Lord
Bingham at [24], Lord Steyn at [49]-[50] and Lord Carswell at [68]-[70] in
particular.
391. We must first of all note that according to the logic of the UNHCR position the ordinary test as applied in the Article 3 context could only be applied in any event if the Eritrean military service system amounts to either slavery or servitude, not if it only amounts to forced labour. That is because for UNHCR the critical factor that determines the test is the status of Article 4(1) as a non-derogable right. But, as the respondent properly highlighted, that right only encompasses slavery and servitude, not forced labour. On UNHCR’s own analysis the test to be applied if the Eritrean military service system amounts to forced or compulsory labour only, would be that of ‘flagrant breach’.
392. At all events, we do not consider that their Lordships intended in Ullah to prescribe a specific legal test for the extraterritorial application of Article 4(1). Indeed their remarks regarding all of the nonderogable rights of the ECHR other than Article 3 were predicated on their understanding being descriptive of what the position was in Strasbourg jurisprudence: see e.g. Lord Bingham at [68] and Lord Steyn at [50]. That is important because the respondent’s position that Article 3 “remains … a special case” would in logic seem only to hold if the Strasbourg Court has excluded that the same test could apply in an Article 4 extraterritorial context.
393. We have not been taken to any Strasbourg or other national case on Article 4 in an extraterritorial case that applies a “flagrant denial” test. On the other hand, the Strasbourg Court seems to have come close to applying much the same test as applied for Article 3 in the case of VF contre France Application no 7196/10, 29 November 2011 where the applicant claimed, inter alia, that her return to Nigeria from where she had been trafficked, would place her at risk of being again forced into prostitution. In the event the Court concluded it was not necessary to pronounce on this issue since it considered the application manifestly ill-founded, but its analysis of why it came to that conclusion made no reference to a “flagrant breach” test and proceeded on the basis that the issue was a straightforward one of real risk of a violation. When it came to the applicant’s further claim that her return would violate Article 3, the Court said that it did not consider it necessary to address it since it had been considered in substance under Article 4.
394. We do not regard as conclusive as to what the test should be for Article 4(1) that the Strasbourg Court has not seen the same test as Article 3 to apply to Articles 2 and 7, which are also non-derogable (a “near certainty test” being applied to Article 2 -see Lord Bingham at [15] of Ullah, citing para 61 of Dehwari (Dehwari v Netherlands (2000) 29 EHRR CD 74) because, again, the premise of any analysis in this regard is what has been established by Strasbourg in its jurisprudence. Nor do we think that it is helpful to place reliance on the fact that even rights that could be said to be more important than Article 3 (e.g. Article 2 and the right to life) apply a higher threshold than real risk of a mere violation.
395. But the principal difficulty we have with the respondent’s approach on this issue is that to which we have already alluded above at [376] namely that there is clearly scope for heavy factual overlap between Articles 3 and 4 such that a real risk of being exposed to slavery or servitude (or indeed forced labour) would very often, other things being equal, constitute a real risk of being exposed to treatment contrary to Article 3. It would be odd if the same set of facts showing that there was a real risk of a person being exposed to slavery or servitude or forced labour could result in a finding of a violation of Article 3 but not of Article 4, by virtue of the latter requiring a higher threshold. This is particularly so because, although derogable, Article 4(2) does not identify permissible limitations but only exceptions.
396. In any event, it must be recalled that in the context of assessing whether there is a risk on return at a general level of persons being exposed to a regime of military service contrary to Article 3 or 4, it is necessary to be satisfied that there is a high likelihood that such risks will arise. This approach is, we adjudge, consistent with that taken by the CJEU in the Shephard case (Andre Lawrence Shepherd v Bundesrepublik Deutschland) (Case C-472/13)).
397. To an analogous issue of whether return would expose a person to having to commit acts contrary to international law. The Court ruled at [40] that:
“the assessment which the national authorities must carry out can be based only on a body of evidence which alone is capable of establishing, in view of the circumstances in question, that the situation of that military service makes it credible that such acts will be committed.”
At [43] it concluded that:
“It follows that, in those circumstances, it is for the person seeking refugee status under Article 9(2)(e) of Directive 2004/83 to establish with sufficient plausibility that his unit carries out operations assigned to it, or has carried them out in the past, in such conditions that it is highly likely that acts such as those referred to in that provision will be committed.”
398. Given the findings we go on to make, we resolve to first examine the Article 4 issues posed by Eritrean military/national service at a factual level and then consider whether it makes any difference to our findings if we apply a real risk of a “flagrant breach” or “real risk” of a breach. That way the respondent cannot complain that we have applied a test less stringent than the one she considers apt.
The Eritrean context: the ILO background
399.
On
a number of occasions the various organs of the ILO, including its Committee of
Experts, have found that the Eritrean system of open-ended compulsory national
service constitutes forced or compulsory labour contrary to the ILO
Conventions, in particular the Forced Labour Convention, 1930 (No.29), ratified
by Eritrea
in 2000. The terms of the prohibition on forced labour set out in
this Convention are not precisely the same as those set out in Article 4(2),
but because the Strasbourg Court has treated it as a relevant source of
interpretation and as a starting point for interpretation of Article 4(2) (Van
der Mussele v Belgium [1983] ECHR 12, 23 November 1983 [para 32]; Graziani-Weiss
v Austria [2011] ECHR 1730, 18 October 2011; Stummer v Austria [GC]
[2011] ECHR 1096, 7 July 2011 [para 118]), the ILO assessments are of
particular import.
400. The various ILO materials which were produced to us, which include the very recent Individual Case (CAS) - Discussion: 2015, Publication 104th ILC Session (2015) (and which records the dialogue between Eritrean government representatives) and the Observation (CEACR) –adopted 2015, published 105th ILC session (2016), can be summarised as follows. It is noted that the Eritrean government and worker and employer representatives continue to voice the same views they had been expressing for a number of years namely that the Eritrean system is compatible with the requirements of the Forced Labour Convention because it falls within the permitted exceptions relating to military character (Article 2(2)(a)); normal civic obligations (Article 2(2)(b)); and cases of emergency (Article 2(2)(d)). In relation to the emergency exception, the Eritrean representatives continue to argue that the ongoing border conflict and the absence of peace and stability has been affecting the labour administration of the country and that the “no peace, no war” policy and their concerns about the “threat of war and famine” justify the forced and compulsory nature of the current system of national service.
401.
These
arguments continue to be strenuously rejected by ILO organs. For example, as
regards the need for compulsory military service to be of a purely military
character, it is observed that that limitation has its corollary in Article
1(b) of the Abolition of Forced Labour Convention 1957 (No.105) which prohibits
the exaction of forced or compulsory labour “as a means of mobilising and using
labour for the purposes of economic development”. The practices adopted by the
government of Eritrea
continue to be considered to go well beyond the context
envisaged by Convention No.29 as they allow conscripts not only to be used for
ordinary public works, but also in the private sector. The work exacted from
recruits as part of national service, including work related to national
development, is not considered to be military in character. As regards the
claim that the Eritrean system falls under the emergency exemption set out in
Article 2(2)(d), ILO organs continue to regard this exception as applying only
in restricted circumstances confined to genuine cases of emergency, or force
majeure, that is, sudden, unforeseen happenings calling for instant
counter-measures. According to the ILO organs, the Eritrean system of national
service, being in force for over two decades, cannot benefit from this
exception.
Slavery and servitude: our assessment
402. The principal basis on which the appellants contend that the Eritrean national service system amounts to slavery is the conclusions of the 2016 UNCOI Report to this effect. We would note that we think they are entirely right to focus on the 2016 Report because the 2015 Report, although containing at Part 6 a section C headed “Abused, Exploited and Enslaved”, only refers glancingly to enslavement or servitude (e.g. at n.2093) or “slave-like” conditions and only illustrates such concepts in the context of the treatment of women in military contexts; it nowhere refers specifically to either international human rights law (IHRL) prohibitions on slavery and servitude (e.g. Article 8, ICCPR) or specific customary international law provisions.
403. We have considerable reservations about the reasoning adopted in the 2016 Report as regards slavery and servitude.
404.
First,
although stating that it is guided, inter alia, by IHRL and customary
international law ([6]), the Commission’s analysis is conducted in the context
of deciding whether the Eritrean system of military/national service amounts to
enslavement as a crime against humanity - as defined by Article 7 of the Rome
Statute (see e.g. [191]) or by equivalent customary international law (e.g.
[196]-[197]). Correspondingly, the jurisprudence it bases itself on is that of
the international criminal tribunals such as the ICTY and the ICC. That the
Commission should choose that context is only to be expected given that its
mandate had been extended for this purpose “in order to investigate systematic,
widespread and gross violations of human rights in Eritrea
with a view to
ensuring full accountability, including where these violations may amount to
crimes against humanity”([(3]). But it does mean that for our purposes we
cannot treat its analysis as being based directly on IHRL - either Article 8 of
the ICCPR or its European equivalent in Article 4 of the ECHR. Our task is
limited to deciding whether the Eritrean system violates Article 4 of the ECHR
and despite more than one opportunity to do so, the ECtHR has not seen the
international criminal law framework as providing guidance for the
interpretation of Article 4.
405. Whilst it has oppressive features, we do not consider that the Eritrean system of military/national service constitutes anything comparable to the paradigm identified in Siliadin of “the obligation for the 'serf' to live on another person's property and the impossibility of altering his condition”, certainly not in the context of assessing the military/national service system as a whole, whose conditions are extremely variable: see above [267], [274] and [288]. Even those who are required to perform lengthy national service cannot sensibly be described as being compelled to live permanently on government property and whilst the possibilities for exemption or de facto demobilisation are limited, it cannot be said that there is an impossibility to alter one’s condition. Nor do we consider that the obligation to perform military/national service can sensibly be described as amounting to the “exercise [by the Eritrean state] of a genuine right of legal ownership …. reducing those called up to the status of an “object””. Eritrean law does not create such a legal ownership.
406. We entirely follow the Commission’s summary of the approach taken in the international criminal law context by the ICTY trial and appeals chambers in Kunarac (Kunarac [2001] IT-96-23-T/IT-96-23/1-T (Tribunal) and Kunarac [2002] IT 96-23 8 IT96 23/1-A (Appeals Chamber)) and that taken by the ICC Trial Chamber in the Katanga case (Kantanga [2009] IT-95-5/18-AR73.3) which consider that the powers attaching to the right of ownership should not be construed as limited to the crime of “chattel slavery” and regards a number of other indicia of ownership and control (ten in total) as being relevant. However, even on the Commission of Inquiry’s own application of these indicia to the Eritrean context, we do not follow how it progresses from its argument that there are certain aspects of the Eritrean system of military/national service that constitute the crime of enslavement to its conclusion that the programme generally, including civilian national service and service in the people’s militia, constitutes such a crime.
407. Of the ten indicia relied on to justify the finding that the system amounts to enslavement, there are at least three that can only be applied to civilian national service and the people’s militia with considerable difficulty: e.g. “(vi) inhumane conditions”, “(vii) torture and killing” (where all the examples cited relate to military national service, not civilian national service) and “(x) impact on family life”. As regards civilian national service, the 2015 UNCOI Report itself, for example, notes at [1443] that:
“[c]onditions in civil service are perceived to be far better than in the army because conscripts may lead a civilian life. They have regular office working hours. Outside working hours, their time is free and they usually have at least part of the weekend off. …Conscripts are free to live with their families, may attend religious services outside of working hours and can get married without restriction or prior authorisation. Some may get annual leave, but others have none. “
408. The Commission goes on to note, however, that freedom of movement of those in civil service is restricted. Having noted at [1446], that conscripts in the army are frequently subjected to punishment in connection with the labour exacted from them that amounts to torture, the Commission observes at [1447] that: “Unlike those in the army, conscripts in civil service are usually not subjected to harsh punishment in the course of their work. When they leave work without authorisation, they are treated differently from conscripts in the army. ..”
409. What the Commission appears to rely on for including even civilian national service within its categorisation of the Eritrean system of military/national service as amounting to slavery is the lack of freedom of choice. Thus the Commission observes at [2010] of its 2016 Report that:
“As noted above, the Commission has heard evidence that some conscripts are assigned to work in non-manual labour in government ministries, schools, hospitals and in the judiciary, but that even these conscripts have no freedom of choice”.
410. There is then a reference back to [90] where it is noted that “[t]he working conditions for this set of conscripts, particularly for those working in Asmara, appear to be more favourable but that “these conscripts have no freedom of choice.” However, we are not aware that lack of freedom of choice (even when coupled with features such as restricted freedom of movement, occasional disproportionate punishment for absenteeism etc), is sufficient to constitute the crime of enslavement or (more pertinent for our purposes) a violation of the Article 4(1) prohibition against slavery.
411. Nor in relation to these indicia, does the Commission’s own description regard them as applicable either at all or to the same extent in respect of the people’s militia.
412. Even applying the international criminal law framework, the Commission appears to make an unjustified leap from the identification of instances where the ten indicia apply to the conclusion in [234] that “…there are reasonable grounds to believe that Eritrean officials have committed the crime of enslavement, a crime against humanity, in a persistent, widespread and systematic manner since no later than 2002.” What is missing from the Commission’s analysis is any concrete basis for considering that the scale of the violations of each of these indicia is such that, quantitatively and qualitatively, it can be said to cross the threshold of “widespread and systematic”. (In this regard, the Commission’s decision in its 2016 Report not to probe the evidence of any of the respondents who sent responses to their first report as to what light, if any, it might shed on the scale and frequency of such violations, does not assist). We remind ourselves, that by operation of Article 7(1) of the Rome Statute this threshold is a necessary condition for there to arise a “crime against humanity” (“For the purpose of this Statute, "crime against humanity" means any of the following acts when committed as part of a widespread or systematic attack directed against any civilian population, with knowledge of the attack”). We also see a difficulty with the Commission’s reasoning for classifying conscripts generically as civilians, but do not consider it necessary to develop this point here.
413. We note further that the case that national service is slavery is not supported by the evidence of PK. His evidence was that some even do it voluntarily. The clear thrust of his evidence was that what is problematic about national service is that it is open-ended and badly paid.
414. We consider that very similar difficulties apply when one turns to consider whether, even if not slavery, the Eritrean system of military/national service amounts to “servitude” contrary to Article 4(1).
415. Having set out our main conclusions on Article 4(1), we turn briefly to consider what difference would be made to them by applying a “flagrant breach” or a “breach” test. Here we would simply observe that we are quite satisfied that the Eritrean system does not pose a “real risk” of a violation of Article 4(1) and a fortiori it could not constitute a real risk of a “flagrant denial” of this provision.
Forced or compulsory labour: our assessment
416. That leaves the issue of whether the system amounts to “forced or compulsory labour”.
417. In this context and in light of the legal framework summarised earlier, it seems to us that the evidence we have before us is on a different footing. For one thing we have the ILO analysis and (unlike the international criminal law framework) the ECtHR has seen the ILO framework to have a bearing on interpretation of Article 4 (see above [399]). For another, the ILO analysis, taken together with other sources, constitutes a considerable body of very specific evidence tending to show that the workings of the Eritrean system cannot be seen to fall under any of the exclusions set out in Article 4(3). That is important because in the course of various ILO proceedings the Eritrean government has not disputed that their military/national service system amounts to forced or compulsory labour. Their argument is directed only to their system falling under one or more of the permitted exemptions or exclusions.
418. We take first the exclusion of ‘any service of a military character” (Article 4(3)(b)).
419. Paragraph 3(b) of the Article excludes from the ambit of the term “forced or compulsory labour”, as used in paragraph (2), “any service of a military character”. There are at least two respects in which the Eritrean system of military/national service falls outwith this exclusion. First of all, its legislative framework, Article 5 of the 1995 decree in particular, identifies one of the objectives of military service as “to develop and enforce the economy of the national by investing in development work….” The legislative framework thereby endorses the use of compulsory labour for purposes of economic development. Second, there is overwhelming evidence that in its actual practice the Eritrean state uses conscript labour for services of a non-military character. The 2015 UNCOI Report documents the use of conscript labour in construction projects and in support of private enterprise, in agriculture, in the civil service and in the judiciary ([1399]-[1479]). In any event we do not understand the arguments of the Eritrean representatives before the ILO to dispute the use of conscript labour in the context of a wide range of public works, although they do dispute (unconvincingly in our view) its use for private enterprise.
420. As regards assignment to civilian national service, it seems to us that, notwithstanding that its conditions are not generally oppressive, that is not a necessary condition for forced labour. Here the UNCOI argument concerning lack of freedom of choice (which we rejected in relation to the slavery/servitude argument) has in our judgement a proper place, since the effect is that those forced to undertake such work are prevented often for lengthy periods from pursuing their own occupations and professions, save for some access to small family-based businesses. According to the Council of Europe/ECtHR study, it is not necessary for forced labour to exist that the condition being experienced be permanent or unlikely to change (para 17).
421. We have not found it easy to decide the issue of whether it is correct to conclude that the Eritrean system of military/national service as a whole constitutes forced labour, given that civilian national service does not ordinarily result in significant punishments and can sometimes amount to little more than attending an office in normal working hours and in the case of older women is sometimes said to be undertaken voluntarily. On balance we consider that the breach is a generic one for several reasons. First, the Eritrean government representatives before the ILO have not sought to argue that civilian national service is other than forced labour (although they dispute whether it falls within permitted exceptions). Second, ILO organs have seen it as generic. Third, even though we are unable to accept the findings of the 2016 UNCOI Report that the Eritrean system constitutes enslavement and servitude, it does particularise aspects that have a strong bearing on the issue of forced labour. Thus the 2015 UNCOI Report notes at [1426] that;
“The length and conditions of work for conscripts, including wages, working hours, place of assignment, leave time and rest days do not per se constitute elements of forced labour. But the open ended nature of national service and the often harsh working and living conditions of conscripts subjected to forced labour have a significant impact on the enjoyment of some rights including safe and healthy working conditions, the right to security, integrity of the person, and the highest attainable standard of physical and mental health”.
422. In the same report at [1519] it is stated that the “Government has unlawfully and consistently been using conscripts and other members of the population, including members of the militia, many beyond retirement age, as forced labourers to construct infrastructure and to pursue the aim of economic development and self-sufficiency of the State, thus indirectly government that has been in power for the past 24 years”. Fourth, even if not performed in oppressive conditions, civilian national service (like service in the people’s militia) nevertheless falls within the description of work “exacted …under the menace of any penalty” and also performed against the will of the person concerned, that is work for which he “has not offered himself voluntarily” (Van Der Mussele, para 34). The fact that some older women undertake it voluntarily, because it pays them something, does not seem to us to be enough to alter its underlying character as forced labour. We would also observe that the Home Office CIGs on National Service appears to acknowledge that there may well be a real issue as regards forced labour in the context of Eritrean national service. At 2.3.46 of the August 2016 version, for example, it is stated (with reference to non-civilian national service) that “[w]here a person is able to demonstrate that as a result of the open-ended nature of their national service they will face a flagrant denial of their right not to be required to perform ‘forced labour’, they will be entitled to a grant of discretionary leave…” (see also 3.1.10 and, as regards women, 11.3.3).
423.
In
relation to the exemption for “any work or service which forms part of normal
civic obligations” (Article 4(3)(d)), we consider that the reasoning of the ILO
organs applies with equal force in the context of Article 4 of the ECHR. We do
not consider that the use of conscripts in civilian national service can escape
the application of Article 4(3) on the basis that they form part of normal
civic obligations. It is true that Strasbourg jurisprudence has seen this
exclusion clause to include compulsory fire service in Baden-Wurttemberg (Karlheinz
Schmidt v Germany judgment pp no. 13580/88, 18 July 1994, para 23);
compulsory jury service such as exists in Malta (in Zarb Adami v Malta, Application
no 17309/02, para 47); an obligation to conduct free medical examinations (Reitmayr
v Austria; and the obligation to participate in the medical emergency
services (Steindel v Germany). However, as the ILO organs have
consistently noted, the range and extent of work conscripts in Eritrea
are required to perform in civilian national service goes well beyond anything that
can be described as the performance of “normal civic obligations”,
(emphasis added). The UNCOI Report of 2015 reinforces the findings of the ILO that
national service is a way of controlling the population. Even though we
consider discharge/release is granted more frequently than has been contended
by the appellants and UNHCR, it remains that for those who have to perform such
duties, the type of work a conscript is expected to do is again arbitrary and
includes agricultural work, working in the mining industry and construction
work. There is evidence of conscripts working for the private benefit of
commanders and of the government lending conscripts to foreign companies (e.g.
PK’s evidence, the EASO Report at 3.5 and the evidence relating to the mining
industry).
424. As regards the exemption based on provision of emergency services (Article 4(3)(c)), we consider that the ILO organs are entirely right in their repeated conclusion that the Eritrean reliance over a lengthy period on this provision goes well beyond the restricted nature of this exemption. The 2015 UNCOI Report reinforces the ILO observations, noting at [1468] in respect of the people’s militia for example, that “[T]he Commission is not aware of any such situation of emergency in the last few years that would have justified the establishment of the People’s Army. In any case, by definition, such situations of emergency are limited in time and compulsory labour cannot be exacted beyond the critical and genuine phase of emergency.”
425. We turn then to consider whether our findings regarding forced or compulsory labour would be any different depending on whether we applied a “real risk” or “flagrant denial” test. We are entirely satisfied that the open-ended duration of national service, coupled with the fact that its duration appears to be prolonged, gives rise to a real risk of a violation. There is a significant body of evidence showing that conscripts will be required to engage in work where the conditions amount to forced labour. There is strong evidence of conscripts working in the agricultural and construction industry in poor conditions. There is the Bisha mine evidence. There is strong evidence of poor conditions and mistreatment during military and some types of civic service. However, despite such evidence, we do not find that such conditions are sufficiently widespread for us to conclude that they amount to forced labour. Not all conscripts are working in conditions that would constitute forced labour. Nevertheless, for reasons given above, we consider that the lack of freedom of choice is sufficient to give rise to a breach. We also think that it amounts to a “flagrant breach” of the right to be protected against forced or compulsory labour, since in our views the Eritrean system effectively extinguishes that right.
426. In this regard we would emphasise again that that unlike qualified rights such as Articles 8 to 11, the ECtHR has not regarded the limitations set out in Article 4(3) as being intended to "limit" the exercise of the right guaranteed by paragraph 2. Taken together with the fact that there has been recognition of a strong factual overlap between Articles 3 and 4, (including in respect of forced or compulsory labour) we do not consider that there is a need to establish extinguishment of content beyond that set out in Article 4(2).
427. For similar reasons we also consider that to the extent that the Eritrean system of military/national service breaches Article 4(2) it is also likely to give rise to a violation of Article 3.
428. We would emphasise, however, that our findings above concern active national service only. If one is a reservist subject to recall, we do not find that the risk of recall is sufficiently likely to amount to a breach of Article 4 (see [297] – [307] above).
429.
We
conclude that the national service regime in Eritrea
does not as a whole
constitute enslavement or servitude contrary to Article 4(1) of the ECHR, but
that it does constitute forced labour under Article 4(3) which is not of a type
permitted under Article 4(3)(a)-(d). A real risk on return of having to perform
military national service duties (including civilian national service but not
with the people’s militia) is likely to constitute a flagrant or a mere breach
of Article 4(3) as well as a breach of Article 3 of the ECHR.
430.
Where it is
specified above that there is a real risk of persecution in the context of
performance of military/national service, it is highly likely that it will be
persecution for a Convention reason based on imputed political opinion. In so
concluding we take into account that the Home Office CIG: Eritrea
: National
(incl.Military) Service, Version 3.0, August 2016 at 2.2.6 considers that
given the Eritrean regime’s economic realpolitik and the widespread emigration
“it is unlikely that
avoiding national service, by itself, is now perceived to be a political act by
the government” (see also 2.2.3 and 3.1.3), but note that the same document
cites the May 2015 EASO Report as stating that “[d]ue to the political and ideological
nature of national service, most sources claim that desertion or draft evasion
may be regarded by the authorities as an expression of political opposition or
treason”. Additionally, in this CIG’s Country Information section addressing
this topic, only one source interviewed by the UK FFM (a UN staff member) is
cited in support of the proposition set out at 2.2.6 and (as UNHCR correctly
pointed out in her submissions regarding this CIG) this UN official does not
directly answer the question of whether deserters are treated as traitors by
the government. We do not consider the recent evidence to demonstrate that the
Eritrean regime has ceased viewing national service in political and ideological
terms. The fact (noted in the AI “Just Deserters” Report) that the Wall Street Journal, whose
correspondent was permitted a media trip to
Eritrea
in September 2015, reported
that the Eritrean government had rejected a $222.7 million plan from the EU to
facilitate the demobilisation of long serving conscripts because “it would
violate the principle that no one is exempt from patriotic duties”, only reinforces us in this view.
Conclusions
431. Our conclusions may be stated thus:
Legal
“Country guidance” is an established term denoting judicial guidance and adoption by the Home Office of terminology apt to confuse this important fact is to be deprecated.
Country guidance
(i) Men aged over 54
(ii) Women aged over 47
(iii) Children aged under five (with some scope for adolescents in family reunification cases
(iv) People exempt from national service on medical grounds
(v) People travelling abroad for medical treatment
(vi) People travelling abroad for studies or for a conference
(vii) Business and sportsmen
(viii) Former freedom fighters (Tegadelti) and their family members
(ix) Authority representatives in leading positions and their family members
(i) A person who is likely to be perceived as a deserter/evader will not be able to avoid exposure to such real risk merely by showing they have paid (or are willing to pay) the diaspora tax and/have signed (or are willing to sign) the letter of regret.
(ii) Even if such a person may avoid punishment in the form of detention and ill-treatment it is likely that he or she will be assigned to perform (further) national service, which, is likely to amount to treatment contrary to Articles 3 and 4 of the ECHR unless he or she falls within one or more of the three limited exceptions set out immediately below in (iii).
(iii) It
remains the case (as in MO) that there are persons likely not to face a
real risk of persecution or serious harm notwithstanding that they left
illegally and will be perceived on return as draft evaders and deserters,
namely: (1) persons whom the regime’s military and political leadership
perceives as having given them valuable service (either in Eritrea
or abroad);
(2) persons who are trusted family members of, or are themselves part of, the
regime’s military or political leadership. A further possible exception,
requiring a more case specific analysis is (3) persons (and their children born
afterwards) who fled (what later became the territory of)
Eritrea
during the War of Independence.
E. ASSESSMENT: THE APPELLANTS
MST
432.
Beyond
the fact that MST is a national of Eritrea
, participated in national military
service at some stage and arrived in the UK on 21 November 2014, the only other
fact we are prepared to accept is his account that his family owned livestock
and grew crops. He was entirely consistent about this matter and as a result we
are able to find that his family enjoys secure economic circumstances. However,
as regards the rest of his account, there are significant credibility problems
arising from his evidence. His representatives conceded that he was not an
impressive witness. He gave evidence before Judge Holmes who recorded his
evidence in detail. There was no challenge to the record of evidence. In a
number of material respects it was at odds with what MST told us.
433.
At
the hearing before us he stated, for the first time, that he had given a false
name and date of birth when he was detained (following having been caught on
the border trying to leave Eritrea
). He did not mention this to Judge Holmes
and he raised it for the first time at the hearing before us. Judge Holmes
observed that MST was released when he was aged 18 and allowed to return to his
family despite the fact that he was at that age due to perform national service
(see [24]) of Judge Holmes’ decision). Judge Holmes recorded at [25] that MST
was pressed on this “and his only explanation was that the authorities did not
know how old he was because he did not tell them, and they had no way of
ascertaining his age”. We believe that MST has fabricated this part of his
evidence in a misconceived attempt to overcome the difficulty in his evidence
as highlighted by Judge Holmes.
434. In his evidence before Judge Holmes, MST stated that he had been issued with an ID card in 2008 when he was aged 19. He told us that he was issued with an ID card in 2007; his inconsistency arose from confusion. It is our view that he has fabricated this part of his evidence in another attempt to overcome the difficulties in his evidence identified by Judge Holmes (namely why he would receive an ID card at the age of 19 which would have identified him as being eligible for compulsory national service from the age of 18 and yet he had not received call-up papers despite the Global Administration Centre being aware of his age).
435. MST’s evidence before us was that he married on 16 January 2011 when his wife was aged 22 and that they stayed in his family home for about one month during the honeymoon period, but afterwards he rarely saw her. However, in his evidence before Judge Holmes he stated that his wife was aged 18 when they married and that he had lived with her for a year after their marriage in his family home. He told Judge Holmes that he then left the family home to live elsewhere, visiting her at her parents’ home from time to time. When this was put to him at the hearing before us he said that because his wife was living with his parents he counted this as living with him. We did not find this to be a credible or adequate explanation for the inconsistency. On the issue of his wife’s age, in an attempt to resolve the discrepancies, he told us that his wife was now aged 23 and she was aged 22 at the date of the hearing before the First-tier Tribunal. Again we do not accept this as a reasonable or credible explanation.
436. MST’s evidence before us relating to his escape from the lorry in Mendefera is inconsistent with what he told Judge Holmes. He told Judge Holmes that he stripped off his military uniform in order to blend into the general population; however, he told us that he was not wearing a uniform at the time. When Mr Rawat put the discrepancy to him he stated that he was in fact wearing a military overall which he removed. In our view this was another example of inconsistent evidence further undermining his credibility.
437.
It
does not assist MST that there are significant discrepancies in his evidence
for which he has not given an adequate explanation. He described himself as
single in the screening interview. In evidence before us he stated that he
believed that what he was being asked was whether his wife was with him in the UK, but this is not a reasonable or credible explanation for describing himself as single
if indeed he is married. In the screening interview he was asked about his
occupation and indicated that he had completed national service (Q1.9). We
have taken into account his explanation before us that this was an error.
However, later in the same interview he was asked (Q5.7) whether he had ever
worked for organisations including the armed forces and he gave the dates
February 2009 to October 2009 for national service. In a Bio-Data Information
form he described himself as unemployed in Eritrea
. He was asked whether he has
ever been arrested and whether he is subject to an arrest warrant or wanted by
any law enforcement authority for an offence in any country (Q5.1 and Q5.2) and
he answered no to both questions. We have considered MST’s evidence that the
answer in relation to national service was erroneously translated. We have
taken into account his evidence that he has not committed an offence and that
is why he answered no to Q5.2, but this is not consistent with the background
evidence in relation to
Eritrea
. Illegal exit and desertion from the army are
both criminal offences and it is inconceivable that he would not be aware of
this.
438. MST was asked (Q7.1) whether he has been subject to any forced work or exploitation in his country and he answered no. In his evidence before us he described poor conditions at Wia, but this is in contrast to what he said in the Asylum Interview at (Q43) when he said that he did not experience problems whilst doing military training.
439. MST failed to put forward a credible or reasonable explanation why he did not claim asylum in France or Italy.
440. Having found that MST wholly lacks credibility (except in relation to his nationality, participation of some sort in national service, arrival in the UK and his family’s secure economic circumstances), we do not accept his account and reject his evidence.
441. Having been found to be wholly lacking in credibility and his account having been rejected, MST cannot be assumed to have left illegally. Failed asylum seekers are not at risk for that reason alone.
442.
In
accordance with our country guidance, we must thus turn to consider whether as
someone who has obviously left Eritrea
on or after August/September 2008,
inferences can be drawn from MST’s health history or level of education or his
skills profile as to whether legal exit on his part was feasible, provided that
such inferences can be drawn in the light of adverse credibility findings. Given
that his date of birth is 18 February 1989 and that he left
Eritrea
circa January 2013, we find it reasonably likely that he served several years in national
service and that during this time he acquired experience or skills making it
feasible for him to qualify for lawful exit.
443. It follows from our guidance that MST would not be perceived on return as someone who has exited illegally and hence he would not be at risk on return.
MYK
444. Whilst it is reasonable to expect a degree of confusion in relation to dates, MYK’s account is littered with inconsistencies in relation to dates and the relative timing of events to such an extent that he has failed to put forward a coherent account. The most serious inconsistencies are the following:
1). In his asylum interview he indicated that his mother had been imprisoned twice and this is entirely at odds with his evidence in his witness statements and oral evidence. There is also internal inconsistency. In the same interview he stated that the second time the authorities came she was not arrested, but threatened, and that she had not been detained twice. However, in oral evidence before us he said the threatening visit when she was not detained was on an earlier occasion to the arrest. There is no mention of this in his witness statements.
2). Whilst it is reasonable to confuse events in 2011 and 2012, the discrepancies are not limited to simple confusion over dates. In the asylum interview he clearly indicated that he had been on leave in 2011 and in 2012. This is entirely at odds with the evidence on which he now seeks to rely.
3). MYK’s evidence relating to detention is inconsistent in terms of dates and duration. This was put to him during the asylum interview whereupon he changed his account stating that he had in fact been detained for a period of seven months (this fitted with the dates that he had given). However, the discrepancy is not properly addressed and inconsistency remains throughout the evidence. In his most recent witness statement he still maintains that he was detained for a period of two months whilst in the witness statement of 20 January 2015 his evidence was that he was detained from September 2012 until March 2013.
4). In oral evidence before us MYK stated that he returned to his military unit in January 2013 and that it was “a memorable date”, but this is entirely at odds with what he said during his interview; namely, that he returned to his military unit in March 2013. This inconsistency was put to him by Mr Rawat and he stated that when he rejoined his unit he was arrested again. This not only inadequately addresses the point, but he had not mentioned previously having been arrested after January 2013.
445.
We
do not accept MYK’s evidence relating to contact with his wife. It is not
credible that he would not have made efforts to contact her since his arrival
in the UK; particularly considering that he said he had only learnt that he had
had a child after his departure. We do not accept that he has not been in
contact with her since fleeing Eritrea
and conclude that he learned about the
birth of his son having made contact with her. We do not accept that he was
told this by other Eritreans.
446.
We
have taken into account what MYK said about the raising of funds in his
interview (see Q263 and Q264). We found that the evidence generally about the
funding of his journey and payment of agents was vague and inadequate. There
was no credible explanation given how funds had been raised in such a short
period of time since receipt of the call-up papers (he left Eritrea
in the same month). The evidence about making contact with his uncle once in Sudan, how he was able to track down the smuggler in Libya after his escape and how he obtained
funding to continue his journey into Italy and then into France and the UK, is similarly vague and unsupported.
447. His evidence before us was that he had lost his identity card, but there was no explanation why this had not previously been mentioned. It is inconsistent with what he said in his screening interview (Q2.6) where he said that he had not ever had his own Eritrean ID card/military identity card or driving licence.
448.
MKY
has failed to provide a coherent or credible account relating to his
circumstances in Eritrea
, the reasons for having left
Eritrea
and the funding of his journey here. It follows that we reject his account.
449. Having been found to be wholly lacking in credibility and his account having been rejected, MYK cannot be assumed to have left illegally. Failed asylum seekers are not at risk for that reason alone.
450.
In
accordance with our country guidance we thus turn to consider whether as
someone who has obviously left Eritrea
on or after August/September 2008,
inferences can be drawn from his health history or level of education or his
skills profile as to whether legal exit on his part was feasible, provided that
such inferences can be drawn in the light of adverse credibility findings. We
think it reasonably likely that he has performed several years of national
service duties whilst in
Eritrea
(and that was his own account) and that during
that time he acquired experience or skills making it feasible he would qualify
for an exit visa.
451. It follows from our guidance that MYK would not be perceived on return as a draft evader or deserter and would not be at risk on return.
AA
452.
AA
did not give evidence before us. He has a chronic mental illness. He is
currently prescribed quetiapine, an antipsychotic drug. He has been discharged
into the community and is in 24 hour supported accommodation where he is
monitored. It is accepted by the respondent that quetiapine is not available in
Eritrea
, but it is the respondent’s case that other antipsychotic medication
is available as is medication to counter side effects. The First-tier Tribunal
found at [71] that there is a family home in
Eritrea
and some reason for
believing that AA’s family do continue to spend time there in addition to
residing in Saudi Arabia. At [86] the First-tier Tribunal found that there are
some “remote” family members in
Eritrea
. There has been no challenge to these
findings and no reason for us to go behind them. It is not entirely clear
whether on the accepted evidence AA could be said to have exited
Eritrea
illegally, but since even on the basis most favourable to the respondent – that he
left lawfully – the findings we make below would still be the same, we shall
assume in what follows that he left lawfully.
453. We have taken into account all the medical evidence relating specifically to AA and the conclusions we reached concerning medical exemptions generally. We take into account the latest correspondence from Dr Larsen concerning alternative medication and medication to counter side effects. There is no proper challenge by the respondent to this evidence and we accept that it establishes that there is a reasonable risk of side effects in the event of a change of prescription. We also accept that there is a strong clinical argument for AA to continue taking quetiapine.
454.
Having
accepted the medical evidence relating to AA we find that when AA is taking his
prescribed medication, quetiapine, he does not present with visible symptoms of
being mentally ill and therefore he would not present as medically unfit on
arrival to Eritrea
. His condition would deteriorate without medication. From
the evidence before us we are not able to say with any certainty when this
would take place, but in the absence of evidence to the contrary, we accept
what Ms Robertson told the First-tier Tribunal that there would be symptoms
within a couple of weeks (see [40] of the decision of the First–tier
Tribunal). AA does not speak fluent Tigrinya. We conclude that it is
reasonably likely, at least in the short-term, that AA will suffer a relapse,
not long after his return.
455. It is AA’s case that the circumstances of his illness reach the threshold required under Article 3 in the context of health cases, and secondly; he will not be exempt from national service; rather he will be required to undertake national service. In the light of his mental health, this would amount to treatment contrary to Article 3 and Article 4. We will engage with the second proposition advanced by AA relating to Article 3.
456.
AA
would not be returning to Eritrea
as an evader or a deserter, but he would be
required to do national service, unless subject to exemption on grounds of
ill-health.
457. We conclude that it is reasonably likely that on arrival AA would be taken to Sawa or a similar military training camp. He would not necessarily be medically assessed. AA would be able to communicate and he is likely to have documents from the UK relating to his condition, but we are not nearly persuaded that it follows it is reasonably likely he would be properly medically assessed and exempted. Whether AA is medically assessed will depend on whether he is able to persuade his commander that he should be. The response may be dependent on whether he starts to exhibit signs of being unwell and this will occur over a period of time, the duration of which is unclear. Throughout the process, it is not reasonably likely that the AA would have access to alternative anti-psychotic medication or indeed any medical care or support.
458.
If
he is medically assessed and it is decided by the medical assessors that he is
unfit, this is not the end of the story because ultimately the decision as to
exemption will lie with his military commander. We note that paragraph 2.3.32
of the Country Information and Guidance: Eritrea
: National (incl. Military)
Service, Version 3.0, August 2016 notes that “a person who is medically unfit and / or disabled, is, depending
on the degree of their impairment, more likely to be assigned to a civilian
post”. However, that statement is immediately qualified
by the words, “… persons have limited choice or
ability to influence where they may be deployed” and we consider
this statement to refer not to initial processing but rather eventual assignment
after being taken to Sawa military training camp. We
are satisfied that there is a reasonable risk of AA having to undergo national
service and initially military training. If he were to be assessed and
ultimately exempted, there is simply no evidence of a clear procedure or
timeframe, which would enable us to conclude that leading up to exemption, and
whilst subject to military training, he would not be at risk of treatment
contrary to Article 3 on account of his mental health.
459. For the above reasons we conclude:
The First-tier Tribunal in the cases of MST, MYK and AA materially erred in law and their decisions have been set aside.
The decisions we re-make are to dismiss the appeals of MST and MYK but to allow the appeal of AA.
Signed
Dr H H Storey
Judge of the Upper Tribunal
APPENDICES
The Evidence of MST
1. At the hearing before us MST gave oral evidence in Tigrinya through an interpreter. He adopted his witness statement of 16 July 2015 as his evidence-in-chief. MST was at the date of the hearing before us aged 27. We will record his oral evidence before that contained in his earlier witness statement because the witness statement is a response to the decision to refuse his application for asylum and is not a detailed account of his evidence.
The AIR
2.
MST married his wife on 16 January 2011 and she is still in Eritrea
. He attempted to flee
Eritrea
in 2006 when he was aged 17 to avoid conscription but
he was arrested on the border at Ghirmayka and detained underground in Adersere
prison for 12 months. He was released without conditions in 2007 and he began
national service in February 2009. He completed military service in October
2009 when he escaped from a lorry in Mendefera. He returned to his village on
foot. He lived in the wilderness for four years, but fled when the authorities
came to his family home and gave his father a final warning. He fled
Eritrea
on 20 January 2013 using an agent and travelled to Sudan, Libya, Italy, France and finally arriving in the UK.
Oral Evidence
3. MST’s oral evidence that he gave at the hearing before us can be summarised. He was born in Adi-Mahkok where he was raised by his parents who are peasant farmers. His family had thirty goats and cattle and grew various crops, but they were not wealthy. They were ordinary. He started school at the age of 11. He dropped out of school, aged 16, with no qualifications, in order to help his parents. His siblings were in the army. Having dropped out of school, he lived with his livestock in what he described as the “wilderness” or “no man’s land.” His nephew would come and tell him when there were roundups and he would then exercise caution. Roundups were seldom in the small rural village where they lived.
4. After
working as a shepherd and subsistence farmer for a year he decided to attempt
to leave the country. He was not free to take his animals to the market in his
home town because he was at risk of being caught by the authorities. He wanted
to leave because he could not have a peaceful life in Eritrea
. His siblings had all been conscripted. They were the property of the State and like slaves.
They were all denied leave to visit their parents. His eldest brother was
married with children and unable to look after them.
5. When he was aged 17 he travelled to Asmara by car and then headed to Sudan on foot. After 20 days he was captured in Ghirmayka, on the border with Sudan, and detained in Adersere Prison for twelve months. He was held in an underground dungeon. The temperature was boiling and conditions generally poor. He was given three pieces of fermented bread to eat each day and dirty water to drink.
6. MST provided the authorities with a false name and false date of birth and misled them into thinking that he was under the age of conscription. Following this he was released in 2007 rather than forced to do national service. He cannot remember whether he was asked about this in his interview with the Home Office. He returned to his village where he continued to hide with his livestock.
7. MST says he started national service on 1 February 2009, at the age of 20. He accounts for this by having given false details when he was detained. He was sent to Wia training camp for eight months where life was miserable, but slightly better than that during his incarceration in Adersere Prison. It was very hot and the food was inadequate. The conscripts were beaten with a stick and put under the sun. He completed military service and was to be posted to a unit.
8. Whilst being transported in a lorry to his unit he managed to escape. He did not know where he was going, but he recognised the area when passing Mendefera and at that point in the journey he escaped by jumping from the lorry. It was market day and it was crowded. He just kept on running and did not look back. He made his way back to his village which was 80 km away. He removed his military overalls. He was wearing light clothes under this and he was given a top and some pants by people he met. He made his way back home on foot, using back routes. The journey took him about two days.
9. MST remained in hiding between 2009 and 2013. He lived with his livestock. He married on 16 January 2011. His wife was aged 22. She was not doing national service. They stayed in his family home for about a month during their honeymoon and afterwards he rarely saw her as he returned to live with his livestock.
10. The authorities were conducting frequent roundups and he decided to flee after his father was given a final warning for him to surrender to the authorities and informed that he would be responsible should his son not surrender.
11. MST has been in contact with his family since he has been in the UK but that was only once and he does not know where his father and uncle are. The authorities did not take action against his father.
12. His paternal uncle arranged for an agent to take him to Sudan and the agent and MST travelled on foot across the Sudanese border. He made his way to Khartoum and from there he travelled on the back of a lorry to Libya where he resided for a year and four months. During most of the time he was in Libya he was in Ganfuda Prison. When celebrating Eid-al-Fitr detention was relaxed and he escaped. He travelled to Italy. This was paid for by his family selling livestock. He continued his journey to France and then to the UK. He did not claim asylum in Italy or France. He was told by the smugglers to continue his journey to the UK.
13. At 1.9 of the screening interview MST was asked about his occupation and his answer is recorded as unemployed/completed national service. His evidence before us was that this must be an error because no one would say that they had completed national service because it is endless. It is also an error when he described himself as single.
14. MST
was given an ID card in Eritrea
in 2007. He denied having told the First-tier
Tribunal that he had been given an ID card in 2008. He had not received call-up
papers. He fears return to
Eritrea
because he has absconded from national
service and left
Eritrea
illegally. He would be subject to imprisonment and he
would also be conscripted and he does not want to be a soldier for the rest of
his life.
Witness Statement of 16 June 2015
15. MST’s witness statement is brief and a response to the decision to refuse to grant him asylum. He said he was not married in the screening interview because he thought that the question was whether his wife was in the UK. The final warning given to his father was verbal. He is wanted by the authorities but he does not know whether there is a warrant for his arrest.
The Evidence of MYK
16. MYK has made three witness statements. The first witness statement is dated 9 August 2014. He was interviewed by the Home Office on 19 November 2014 and the respondent relies on the Asylum Interview Record (AIR). He has produced two further witness statements, of 20 January 2015 and 7 June 2016. He gave oral evidence before and he adopted his witness statements as evidence-in –chief.
Witness Statement of 9 August 2014
17.
MYK was born in Segenetiy, Eritrea
. He married his wife in an arranged
marriage on 15 January 2012 whilst on leave. He last saw her in January 2014
and they have one child, a son, born in March 2014. He left
Eritrea
on 11 January 2014. He attended school between 1998 and 2006. In February 2006, he was
rounded-up by the authorities whilst at school and taken to Wia Military
training camp where he stayed for three weeks. He was then dispersed to Assab
where he remained for nine months. The training there was very tough and
exhausting. He was then dispersed to Gelalo in around December 2006 where he
remained for about a year before being re-deployed to the Gash Barka region
where he was a guard. In December 2010 he was relocated to Tokombya.
18.
He was given home leave for a month in January 2012 and he did not return. He
went into hiding for six or seven months. The authorities came searching for
him in August 2012 and they arrested his mother and she was placed in detention
for about a month. As a result of his mother’s detention, in September 2012, he
handed himself in and his mother was released the same month. He was detained
in Alebu detention centre for two months where he was detained in appalling
conditions and mistreated. After two months he was released and re-joined his
unit. He requested home leave in 2013 and this was refused. He left in May 2013
without permission. He returned home and went into hiding. During this time the
authorities came to his house to look for him. They did not arrest his mother
because she was unwell. He received a call up letter in January 2014 which gave
him until later that month to hand himself in. He then decided to flee Eritrea
. He left the family home on 6 January 2014 with the help on an agent. This was
arranged and paid for by his family. He arrived in Sudan on 11 January 2014. He
arrived in Khartoum on 12 January 2014 where he remained until 18 March 2014
living under the constant fear of deportation. It was common for the Eritrean
authorities to come to Sudan and with the Sudanese authorities to deport
Eritreans. He left Sudan with an agent who was paid by his uncle (who lived in
Eritrea
). He travelled to Libya by lorry through the Sahara desert arriving
in Libya on 5 April 2014. In Libya he was detained in appalling conditions in
a prison near to Tripoli for approximately three months. He was regularly
beaten and forced to carry out hard labour. On 6 July 2014 he managed to
escape. He was with a few prisoners who scattered in different directions and
it was at this point that he took the opportunity to flee. He managed to avoid
being shot by the Libyan authorities and made good his escape. He caught up
with two other escapees and was able to make contact with the agent who had
taken him to Libya. The agent took them to his home in Tripoli where he stayed
for a week before leaving Libya on 13 July 2014. His friends in Israel paid the agent who then took him to Italy. He arrived in Sicily on 14 July 2014.
19.
He was not fingerprinted by the authorities in Italy. He was taken to a
compound where he was able to escape after two hours of arrival. He made his
way to Catania where he met an agent who took him and three others to Rome where he stayed in hiding. He did not make a claim for asylum there. He did not feel
safe in Italy having seen Eritreans living on the street. On 17 July 2014 he
left Italy arriving in France the next day and travelling to Calais where he
stayed (in “the jungle”) until 23 July 2014. Friends paid the costs of the
agent to take him to France. He did not claim asylum in France because he did not have the opportunity to do so and he did not know how to claim asylum
there. He was advised that should he claim asylum in France the authorities would deport him to Eritrea
. With others he managed to travel to the UK in the back of a lorry on 23 July 2014, arriving on the same day. He was arrested in Derby on the same day whilst still in the lorry and he claimed asylum.
The AIR
20. MYK was interviewed by the respondent on 19 November 2014. He stated that he went on home leave in 2011 and when it was put to him that in his witness statement he said he went on leave in August 2012 he stated (at Q130) that he “went on both occasions”. He then said that he went on leave in 2011 and stayed for seven months in August and did not return (see Q132). When he was asked why he said that he had left on two occasions he said that what he meant was that he stayed and that he was still at home in 2012 (see Q132 and Q133). He stated that it might be a mistake in his witness statement that he had been given leave in 2012 (see Q134). He stated during the same interview that he left military service in January 2012 and did not return for seven months and that he had been confused when he said that he went on leave in August 2011 (see Q153).
21. When he did not return to his military unit he was sent a call-up letter and his house was searched (see Q135 and Q136). He received in total two call-up letters (the first when he failed to return to his unit on the first occasion and the second when he left without permission). The authorities came looking for him on two occasions (see Q139). He returned to the military unit in March 2012 (see Q143) in response to his mother’s arrest. His mother was imprisoned on two occasions (see Q148). They took her on the first occasion in December 2012 (see Q149). He later changed this to August 2012 (see Q150). He stated (see Q202), having been asked about the second time that his mother had been detained, that it was after they sent the second call-up letter. He then stated that it was at the end of 2013 (see Q225) but that this time she had received a letter and they did not take her and that she has not been detained twice. The authorities had put pressure on her on the second occasion (see Q228). He then stated that he had left the country and did not know what had happened after that.
22. MYK stated that he returned to his military unit in March 2013. It was put to him in his interview that in his witness statement he had indicated that he had been detained for a period of two months and that if he had handed himself in in August 2012, as he asserted was the case, and was then detained for two months, this would leave a number of months unaccounted for, if he did not re-join his unit in March 2013. In response to this he stated that he was initially working as a prisoner on a farm for two months and then for two months as a normal worker on a farm. After this he had escaped and stayed at home for eight months. He then stated that he had stayed on the farm in detention until March 2013 (see Q190) and had been in detention for a period of seven months. He served with his unit for seven months before escaping in May 2013.
The Witness Statement of 20 January 2015
23. In the witness statement of 20 January 2015, he stated that he was given leave on one occasion in January 2012 and that he did not have leave in 2011. He was sent a call up letter in 2012 which he had forgotten to mention previously. He was detained in Alebu detention centre for five months from September 2012 until March 2013.
The Witness Statement of 7 June 2016
24.
In the witness statement of 7 June 2016, MYK’s evidence is that he is not
prepared to sign a letter of apology and admit that he left Eritrea
illegally having deserted. He is unable and unwilling to pay the 2 per cent tax.
Oral Evidence before the UT on 7 June 2016
25. At the hearing before us MYK was aged 31. He gave oral evidence in Tigrinya through an interpreter. He adopted three witness statements as his evidence-in-chief.
26. MYK
said he had had an Eritrean ID card, but he lost this in the desert en route
to Libya. He has not been able to contact his wife in Eritrea
. He tried to make contact with her through friends, but to no avail. He has been informed by
Eritreans who have met her about the birth of his son. He has not attempted to
contact her since he has been in the UK. She does not have a mobile phone. His
wife did not do military service after they married.
27. MYK was given leave from his unit in 2012, but he did not return. He was on the run for six or seven months before handing himself in. He handed himself in in December 2012 following his mother’s arrest which came about because he had not returned to the military unit. This was the only time that she was arrested. In his asylum interview he made reference to a time when soldiers came to the house and threatened her, but she was not arrested on that occasion.
28.
Having handed himself in, he was detained for two months and then he returned
to his unit in January 2013. When the discrepancy in dates was put to him, MYK
said that he was moved from prison to prison and cannot remember how it
happened although he remembered returning to his unit because it was a
“memorable day”. It was put to him that he had in his interview stated that he
returned to his unit in March 2013 and he then stated to us that when he
returned to his unit he was arrested again and he was not thinking clearly in
his interview. He served for about eight months before he escaped. He received
call up papers in January 2014 before he left Eritrea
. The papers were given to
his mother. As a result of this he decided to flee, but it had always been his
intention to leave
Eritrea
. He left on 6 January 2014 with an agent and fled to
Sudan. He did not know how his family raised the money in order to pay an
agent to help him leave
Eritrea
or how they managed to do this in such a short
period of time (since receipt of the call-up letter in the same month). He
left
Eritrea
illegally on 11 January 2014.
29.
MYK
said he had fled Sudan on 18 March 2014 with the assistance of an agent and
travelled to Libya. His maternal uncle in Eritrea
paid an agent to make the
arrangements. MYK did not have his uncle’s telephone number with him, but he
was able to make telephone contact with him with the help of Eritrean friends
living in Sudan. He was able to contact his friends in Israel with the help of Eritreans and agents. Other Eritreans helped him to pay the agent to take
him to Rome. He did not want to claim asylum in Italy having seen Eritreans
sleeping rough there.
The Evidence of AA
30. AA relied on the evidence that was before the First-Tier Tribunal. He produced a witness statement of 22 September 2014 and he gave oral evidence before the First-tier Tribunal. He also relied on evidence relating to his mental health. AA relied on a witness statement from Ann Robinson, a Deputy Manager at Chalkhill St. Martin of Tours where the appellant was at that time residing and Ms Robinson attended the hearing before the First-tier Tribunal. We do not have a copy of her witness statement, but her evidence has been recorded in the decision. There was before the First-Tier Tribunal a report from Consultant Clinical Psychologist, Mr Nicholas Stokes of 12 August 2013; a letter from Consultant Forensic Psychiatrist, Dr Francis Fernandes of 27 August 2013 and a psychiatric report prepared by Dr John Jacques of 9 September 2014.
31. AA has produced more up-to-date evidence about his mental health. There are three letters from Dr N Larsen, a Locum Consultant Forensic Psychiatrist; the first of 27 April 2015 to Ziadies solicitors; the second of 24 March 2016 to Roelens solicitors and the third of 27 May 2016 to Roelens solicitors. There is a report written by AA’s key worker, Adebisi Ayoade of 6 May 2015, a letter from the Eritrean Community in Lambeth of 9 September 2014 and an expert report from Dr S A Bekalo of 20 April 2015.
32. AA’s
evidence is contained in his witness statement of 22 September 2014. There is
no challenge to the evidence as set out by the First-tier Tribunal in their
decision. AA’s father was born in Eritrea
but his origins and tribal roots are
in Ethiopia. His mother is from an Eritrean tribe. They fled
Eritrea
. AA was born in Sudan in a non-military camp. His family went to Saudi Arabia when he was
three months old and he was raised there. He came to the UK with the help of an agent.
33.
AA’s evidence was that he had returned to Eritrea
on one occasion in 1992,
after the War of Independence, and stayed there for two or three months. He
found the experience frightening. His family reside in Riyadh and his older
sister lives in Canada. He has a cousin and an uncle living in the UK. They did not attend the hearing before the First-tier Tribunal. He speaks to his
parents every four or five months and he misses his family. He can speak and
understand Tigrinya although he does not have a good accent and he cannot read
the language. He would be able to say a few words in the Tigrinyan language
about his diagnosis and how long he has been in hospital. He has no family or
friends in
Eritrea
. His mother has relatives there, however her brothers and
sisters have married and left the country. AA believes that the house
belonging to his mother’s parents is still there. Should he return to
Eritrea
he does not believe that his relatives would be able to help him settle there and
his parents are elderly and would not be able to help him.
34. AA was referred to a community health mental team in 2005. He was detained at the Three Bridges Mental Health Unit between 2007 and January 2014 (following conviction). Since January 2014 he has been at Chalkhill where he receives nursing support and he is able to go out into the community. He sees a doctor every six weeks as well as his community psychiatric nurse (CPN).
The Evidence of Ann Robertson
35. Ms Robertson’s unchallenged evidence is set out in the decision of the First- tier Tribunal. She is a qualified nurse with forensic experience and general mental health experience. She sees AA in the morning and administers his medication. He requires a high level of input from staff. His condition has improved and he engages more and displays less anxiety. There were difficulties regarding his medication. He suffered horrendous side effects from the drug risperidone. He would get up in the morning and then be “wiped out” and would not engage with anybody. His face and tongue were affected.
36. So
long as AA is monitored at the current level, whilst there is always a risk, if
things continue as they are, Ms Robertson has no concerns. If his appeal is
dismissed Ms Robertson would not be surprised if he suffers a relapse and
should he be removed there would be a quick deterioration in his mental health
and stability. Should he return to Eritrea
, with no medication, Ms Robertson
finds it hard to predict what the result would be, but believes that within a
couple of weeks there would be symptoms and AA would deteriorate quickly. He
requires a lot of support.
The Evidence of Dr N Larsen
37. Dr
Larsen’s letters were not before the First-tier Tribunal. In his first letter
of 27 April 2015 he confirms that support and monitoring provided in the 24
hour supported accommodation is essential to AA’s ongoing stability and that he
is at high risk of relapse into acute schizophrenia with significant risks to
himself and others should his mental state be de-stabilised by significant
stressors and that deportation to Eritrea
would be a significant stressor. In
his second letter of 24 March 2016, having been requested to give an update, he
confirms that AA requires long-term treatment with anti-psychotics due to the
chronic nature of his mental illness and the risk of relapse should he
discontinue medication. If he relapses Dr Larsen would expect a return of the
previous symptoms.
38.
In the most recent correspondence, Dr Larsen considers the respondent’s
position that a number of antipsychotic drugs are available in Eritrea
; namely haloperidol, chlorpromazine and fluphenazine. Dr Larsen confirms that AA
has not been previously prescribed any of the drugs purported to be available,
but that he learnt from AA that he has previously been prescribed olanzapine,
risperidone and amisulpride which are all antipsychotic drugs. AA informed him
that olanzapine was discontinued due to excessive weight gain, risperidone was
discontinued due to troublesome side effects and amisulpride was discontinued
due to lack of clinical progress. Haloperidol is not uncommonly prescribed in
psychiatric wards, particularly in cases of acute agitation which are not
uncommon on initial admission to a psychiatric service and it is possible that
AA has been previously prescribed this.
39.
All antipsychotic medications have possible side effects. AA has a history
of sensitivity to the so called extrapyramidal side effects (EPSEs) of
risperidone. Although medication to counter side effects can be prescribed they
are of variable efficacy and Dr Larsen is unsure whether they would be
available in Eritrea
. Fluphenazine and haloperidol result in pronounced EPSEs
although the effect on AA cannot be accurately predicted. Chlorpromazine is
noted to be of moderate potential to cause EPSEs, but a well–established side
effect is sensitivity to light. Dr Larsen would have reservations about
prescribing any of these drugs to AA without clinical trial. The therapeutic
efficacy for switching drugs is difficult to predict. One needs a strong clinical
argument and robust risk management plan in place. It is very likely that AA
would experience side effects from the three drugs and there is a strong
clinical argument for continuing quetiapine in his case.
The Evidence of Dr John Jacques
40. Dr John Jacques was instructed by Ziadies Solicitors in order to prepare an independent psychiatric report on AA. In his report of 9 September 2014 he concluded the following:
(1) AA has a good insight into his mental health problems and past drug use.
(2) AA’s mental state has been stable and has started to improve.
(3) Expulsion would interfere with his mental health. It is most likely that the stress of immigration proceedings has contributed towards his anxiety symptoms which have caused concerns.
(4) AA receives treatment and support in the community and his most recent problems have required a higher level of input from mental health and support services.
(5) AA receives support from a consultant forensic psychiatrist, community psychiatric nurse and vocational support worker. He has a key worker allocated to him at the 24 hour specialist mental health hostel where he lives.
(6)
There is concern that if AA is deported to Eritrea
he would suffer a
relapse. His relapse signature is one of rapid deterioration in his mental
state and behaviour potentially leading to violence as well as self-neglect.
It is therefore important that he receives treatment.
(7)
AA would not receive the same treatment or support in Eritrea
.
Eritrea
does have access to some antipsychotic treatments and basic community support but
the medication he receives, quetiapine, is, according to the Home Office, not
available in
Eritrea
.
(8) AA developed side effects from the alternative drug, risperidone, which included persistent orofacial dyskinesia and hyperprolactinaemia when he was prescribed this medication in 2009. These conditions can be disabling and distressing and there is a risk of permanent irreversible movement problems if the problem is not addressed by stopping the offending medication. Hyperprolactinaemia can have significant problems particularly for men with sexual side effects (breast development and lactation). It is not clear whether AA developed any side effects in relation to hyperprolactinaemia but records indicate that he developed involuntary twitching of the face and tongue which led to discontinuation of the treatment and subsequent resolution of these symptoms.
(9)
The respondent makes reference to other antipsychotic treatments being
available in Eritrea
but such treatments are widely known to cause hormonal
problems and may be less effective in preventing relapse and schizophrenia.
(10)
AA’s death may be expedited if he were unable to access the treatment.
Individuals with schizophrenia have a significantly higher risk of suicide
compared to the general population. His mental health would be at risk of
deterioration if he were to leave the United Kingdom for Eritrea
because he would be unable to access the treatment and support he requires and he will find
adjusting to a new country very difficult without the support from friends and
family and he is fearful of being attacked.
The Evidence of Dr Bekalu
41. In summary Dr Bekalu concluded that if the appellant is returned, the authorities will know that he has not completed national service, he is highly likely to face serious risk on return and he would most likely be subjected to limitless national service.
The Evidence of Adebisi Ayoade
42. Mr Ayoade is AA’s key worker and he provided a report of 6 May 2015 which confirms that AA is currently on quetiapine and is prescribed 400 mgs in the morning and evening and that he has been self-medicating for a month on his evening medication only.
The Background Evidence of Mental Health Problems in the Context of National Service
Amnesty International report of 22 September 2015
43.
In so far as this report engages with the issues that arise in AA’s case, it is
asserted that there is no functioning and reliable process of assessing medical
fitness for national service. The Sawa training facility did not have a routine
health assessment on arrival or at any time. The same is true for other camps.
Permission to see a doctor or a designated first aid officer must be granted by
a commanding officer and it is extremely difficult to obtain. If assessed,
resources are very limited. It is difficult to speak with a great degree of
certainty about the likely outcome for AA if deported because his situation is
unusual. However, anyone forcibly returned from Europe to Eritrea
potentially faces distrust and suspicion and there is a risk that he will be treated as
someone who has tried to flee, the consequences of which are forcible
conscription or arbitrary and indefinite detention. He is unlikely to be able
to obtain medical exemption as he lacks experience and the contacts necessary.
Recognition of mental health difficulties is harder to obtain than physical
health problems. In any event, such exemptions are sometimes ignored in round-ups
or call ups. If conscripted it would be for an indefinite period and constitute
forced labour.
UNCOI 2015
44. The report of 4 June 2015 at [60] concludes that being exempt from national service is very difficult particularly for men. Persons with disabilities are conscripted for military instead of civilian service.
45. Examples are cited within the report of 5 June 2015 of witnesses (with physical injuries) who had not been exempted and forced to remain in military service despite having been declared unfit (see [1196]). The Commission concluded that the exemptions on health grounds are rarely granted, even though the state of health of the persons concerned prevents them from serving in the military. There is evidence of blind and seriously visually impaired people being sent to Sawa (see [1197]).
UNCOI 2016
46. There is an example given within the report at [92] of a witness who in 2014 was unwell with papers to establish this, but who was not believed. The witness reported being detained for six months without due process.
Just Deserters December 2015
47. Amnesty documented (at page 28) reports from former conscripts who told of people with disabilities being conscripted and taken to Sawa for military training. There is no health check or assessment of physical or mental fitness when people are first conscripted and sent for training or at the end of the year at Sawa. Medical assessments are carried out on an ad hoc basis, and usually only if the conscript repeatedly requests it. To obtain an exemption a doctor has to recommend that the conscript is unfit to serve, whether for physical or mental health reasons and this recommendation has to be confirmed by a military commander. Those with health problems have been assigned to national service and the report makes reference to a former conscript with a (physical) health problem who spent three years in national service and another former conscript who had severe injuries to both legs following a car accident. Although the commander concluded that he could not carry out physically demanding tasks, it was decided that he could work. The source stated that this is not a medical decision, but a decision of the Commander. The individual was assigned to administrative work, but he was told there was no pain relief and not granted permission to see a doctor.
The appellants also submitted a transcript of the testimony of Helen Gebreklak dated 21 May 2016 with a certificate of translation. Her evidence touched upon, inter alia, the delays persons in custody face in having their medical complaints dealt with.
Upper Tribunal
(Immigration and Asylum Chamber) Appeal Number: AA/11206/2014
THE IMMIGRATION ACTS
Heard at Stoke |
on 7th May 2015 |
Before
UPPER TRIBUNAL JUDGE HANSON
Between
[MYK]
Appellant
And
THE SECRETARY OF STATE FOR THE HOME DEPARTMENT
Respondent
Representation:
For the Appellant: Mr Howard of Fountain Solicitors.
For the Respondent: Mr McVeety – Senior Home Office Presenting Officer.
ERROR OF LAW
1. This is an appeal against a determination of First-tier Tribunal Judge C Mather promulgated on the 17th February 2015.
© CROWN COPYRIGHT 2016
2.
The
Appellant is a national of Eritrea
born on the 15th August 1987. He
claimed to have left
Eritrea
on 6th January 2014 and travelled via Sudan, Libya, Italy, and France, before arriving in the UK. He claimed asylum and was interviewed on
19th November 2014 and the claim was refused.
3. The Judge found the Appellant not to be a credible witness. In paragraph 17 of the determination it is recorded that the Appellant was unable to give consistent evidence which is said to go to the core of his claim for the reason set out in the refusal letter. No other reasons have been provided.
4. In paragraph 18 the Judge found it not credible that the Appellant did not know how much his mother paid the agent or the agent’s name, but gave no reasons.
5. In paragraph 19 the Judge refers to conflicting evidence by reference to paragraphs 15(i) and (v) of the determination but fails to give reasons and it is not clear what conflicts are being referred too.
6. In paragraph 20 the Judge found the Appellants evidence incredible in relation to how he was able to contact various agents and friends “who magically paid agents for the Appellant during his journey”. The terminology used perhaps reflects the lack of acceptance of the Appellant’s account in the mind of the Judge but no reasons have been given in support of this finding.
7. At paragraph 23 the Judge accepts the Appellant’s nationality and that he undertook his national service. His date of birth of 15th August 1987 is not disputed making him 27 and within the age range of those eligible to serve in the military. At paragraph 24 the Judge states:
“24. I have reminded
myself of the case of MO (Illegal exit – risk on return) Eritrea
CG [2011] UKUT 190, which states:
“c. The general position concerning illegal exit remains as expressed in MA, namely that illegal exit by a person of or approaching draft age and not medically unfit cannot be assumed if they have been found to be wholly incredible.””
8. At paragraph 25 the Judge finds:
“25. I accept the
Respondents’ submissions that the material aspects of the Appellant’s claim are
not credible and I do not accept the Appellant left Eritrea
illegally.”
9. There is no mention of the submissions in the determination. It is not clear on what basis this finding is being made. Are these oral submissions made at the hearing or those in the refusal letter, or both?
10. It is a settled principle that whilst there is no obligation upon a judge to set out their reasons for each and every element of a case before them, it is necessary for the First-tier Tribunal to identify and resolve key conflicts in the evidence and explain in clear and brief terms their reasons so the parties understand why they have won or lost – Budhathoki (reasons for decision) [2014] UKUT 341 (IAC) refers.
11. The Judge may consider that sufficient reasons have been given by referring to the Respondent’s refusal letter and submissions but a reader of the determination is unable to understand the basis of the decision as there is no reference, even in summary form, to the nature of such arguments. It is not clear if the submissions made and issues raised differ in any way from the arguments contained it the refusal letter.
12.
Of
greater concern, which Mr McVeety accepted, is paragraph 24. The quotation from
MO (illegal exit – risk on return) Eritrea
CG [2011] UKUT 00190 (IAC) is arguably selective as the Tribunal also
found that “whilst it also remains the position that
failed asylum seekers as such are not generally at real risk of persecution or
serious harm on return, on present evidence the great majority of such persons
are likely to be perceived as having left illegally and this fact, save for
very limited exceptions, will mean that on return they face a real risk of
persecution or serious harm”. The determination is silent in relation to the
assessment of risk of the Appellant being arbitrarily arrested and ill-treated
on return for this reasons which does no appear to have been considered. The
grounds refer to the Respondents OGN at page 300 of the appeal bundle
indicating this was a matter raised before the Judge.
13.
The reasons for refusal letter relies at
paragraphs 52-55 upon MA (Draft evaders – illegal departures –
risk) Eritrea
CG [2007] UKAIT 00059 and the finding the
issue was that of illegal exit which it is said was upheld in MO (illegal exit – risk on return)
Eritrea
CG [2011] UKUT 00190 (IAC). If the Judge relied upon this
assertion, which may be the case as it appears in the refusal letter, it
ignores the fact that although the Tribunal in MO endorsed the general
position adopted in MA, that a person of or approaching draft age (i.e.
aged 8 or over and still not above the upper age limits for military service,
being under 54 for men and under 47 for women) and not medically unfit who is
accepted as having left
Eritrea
illegally is reasonably likely to be regarded
with serious hostility on return, it found this was subject to limited
exceptions in respect of (1) persons whom the regime’s military and political
leadership perceives as having given them valuable service (either in
Eritrea
or abroad); (2) persons who are trusted family members of, or are themselves
part of, the regime’s military or political leadership. A further possible
exception, requiring a more case-specific analysis, is (3) persons (and their
children born afterwards) who fled (what later became the territory of)
Eritrea
during the war of independence and (v) Whilst it also remains the position that
failed asylum seekers as such are not generally at real risk of persecution or
serious harm on return, on present evidence the great majority of such persons
are likely to be perceived as having left illegally and this fact, save for
very limited exceptions, will mean that on return they face a real risk of persecution
or serious harm.
14. It may be that the Judge was correct in relation to the core of the claim but the lack of reasoning and failure to consider a material element amounts to an arguable material legal error such that the determination shall be set aside. The nationality, fact of having completed national service and immigration history does not appear to be in dispute between the parties but all other elements remain at large.
Decision
15. The First-tier Tribunal Judge materially erred in law. I set aside the decision of the original Judge.
16. Further directions shall follow in relation to the future conduct of this case.
Anonymity.
17. The First-tier Tribunal did not make an order pursuant to rule 45(4)(i) of the Asylum and Immigration Tribunal (Procedure) Rules 2005.
I make no such order pursuant to rule 14 of the Tribunal Procedure (Upper Tribunal) Rules 2008).
Signed……………………………………………….
Upper Tribunal Judge Hanson
Dated the 20th May 2014
Upper Tribunal
(Immigration and Asylum Chamber) Appeal Number: AA/07733/2015
THE IMMIGRATION ACTS
Heard at Field House |
Decision & Reasons Promulgated |
On 15 March 2016 |
|
|
………………………………… |
Before
UPPER TRIBUNAL JUDGE McWILLIAM
Between
MST
Appellant
and
THE SECRETARY OF STATE FOR THE HOME DEPARTMENT
Respondent
Representation:
For the Appellant: Mr G Denholm, Counsel instructed by Immigration Advice Service
For the Respondent: Mr B Rawat, Counsel instructed by the Government Legal Department
© CROWN COPYRIGHT 2016
DECISION AND REASONS
1.
The appellant is a citizen of Eritrea
and his date of birth is 18
February 1989. He made an application for asylum and this was refused by the
respondent on 29 April 2015. The appellant appealed and his appeal was
dismissed by First-tier Tribunal Judge J M Holmes in a determination that was
promulgated on 6 July 2015 following a hearing on 24 June 2015. Permission to
appeal was granted to the appellant by Judge of the Upper Tribunal Jordan on 4 August 2015. Thus the matter came before me.
The Background Evidence
2. The appellant’s case is that he is a deserter and
left Eritrea
illegally. He relied on the case of MO
(illegal exit - risk on return)
Eritrea
CG
[2011] UKUT 190.
3. The respondent’s case is that the appellant is not
credible and, in any event, the guidance in MO should not be
followed in the light of the two Country Information Reports of March 2015,
one entitled Country Information and Guidance – Eritrea
: National (incl.
Military) Service (March 2015) and Country Information and Guidance –
Eritrea
:
Illegal Exit (March 2015).
4. The Danish Immigration Service (“DIS”) produced a report, the Danish Fact-Finding Mission Report (“the FFM report”) which was published in November 2014 and recorded observations on penalties for illegal exit and likely treatment on return. The sources in the report are not identified by name, save Professor Kibreab, but instead are referred to as international organisation or western embassy A, B and C etc. There is also reference to an unnamed well-known intellectual. Reliance was placed on this report by the Secretary of State and informed the Country Information Reports of 2015.
5. It is not necessary for me to set out the guidance
of the Upper Tribunal (“UT”) in MO. Suffice to say, for the
purposes of this decision, that had the appellant established desertion and or
illegal exit (or that he would be perceived as a deserter or as someone who had
exited illegally) and the guidance in MO had been followed, the only
lawful conclusion would have been to allow the appeal. However, the
conclusions of the FFM report are that one who illegally exits Eritrea
/a draft
evader/deserter who has paid 2 per cent income tax and signed an apology letter
would not face problems on return and that the authorities have become more
relaxed and understanding towards the young people who have left
Eritrea
.
Professor Kibreab has distanced himself from the report since publication and
has criticised the findings therein. It has been publicly criticised by others
including the UN, UNHCR and HRW. Professor Kibreab’s position has been
disclosed by a number of emails between him and the DIS. DIS has removed
reference to him in the updated FFM report in December 2014.
6. I was not assisted at the hearing before me by the appellant’s solicitors who had failed to prepare a bundle for the error of law hearing. I was keen to ascertain exactly what evidence was before Judge Holmes. Mr Denholm was not able to assist me. The respondent prepared a bundle for the hearing which purported to include a copy of the appellant’s bundle that was before the First-tier Tribunal (“FtT”) and this includes an index. Mr Rawat indicated that this appellant’s bundle had been served on the respondent, but he had no personal knowledge of what was before the judge.
7. I accept that the bundle replicates that before the First-tier Tribunal, it is apparent that the Judge had before him documents including the following:
1. A document
entitled “Statement on EU Asylum and Aid Policy to Eritrea
of 31 March 2015”.
This document is signed by various academics including Professor Kibreab and
the authors indicate that the FFM report has been the source of
much controversy in Denmark after Professor Kibreab declared that he had been
misquoted and that although the report has not been officially withdrawn its
conclusions are no longer used as a reference for policy in Denmark.
2. A document from
Human Rights Concern Eritrea
expressing concern about the findings of the FFM.
3. A report from HRW
dated 17 December 2014 entitled “Denmark: Eritrea
Immigration Report Deeply
Flawed - European Governments Should Rely on UN Reports, Support UN Inquiry”.
It is asserted that the FFM report is largely based on interviews with
anonymous diplomatic and other sources in
Eritrea
and contains contradictory
and speculative statements about
Eritrea
’s human rights situation. It is
asserted that the sources often qualify their statements about
Eritrea
’s human
rights noting that there is no independent access to detention centres and that
the fate of people returned to
Eritrea
is unclear, but this is not reflected in
the conclusions of the FFM. It is asserted that there is no indication that
the authors of the report interviewed victims or witnesses of human rights
violations in
Eritrea
and a prominent Eritrean academic consulted for the
report has publicly criticised it.
4. A press release from DIS of 9 December 2014 documenting communication between them and Professor Kibreab. It is stated that DIS received an email from Professor Kibreab in which he expressed objections to the report. Corrections and additions were made following this. On Tuesday 25 November 2014 the report was published and a copy sent to Professor Kibreab who sent DIS an email in which he expressed his gratitude for a well-written and informative report.
On Friday 28 November 2014 DIS received an email from Professor Kibreab in which he expressed objections. On the same day DIS received a copy of an email from Professor Kibreab addressed to a number of professionals in which he claimed that DIS attributed information to him which was taken out of context.
The same day DIS asked Professor Kibreab to forward to them his objections but he did not responded to this.
5. A newspaper
article of 10 December 2014 entitled “Denmark admits ‘doubts’ about Eritrea
report” and in this document it is reported that DIS has been under heavy fire
since the report’s release and DIS now says that the feedback “raises doubts”
and Eritreans can expect to be granted asylum in many cases.
It is also
stated that DIS has changed its mind about the conclusions of the much
criticised report after the report was criticised, by its only named source,
Professor Kibreab. It is stated that according to DIS sending deserters of Eritrea
’s compulsory military service back home does present a danger after all and the article
states that in a press release DIS stated that the reaction to its report
“raises doubts about whether there are risks to people returning to
Eritrea
after illegally leaving the country and avoiding national service”.
6. A document from UNHCR in which examples are given of where the FFM report ascribes statements to interlocutors that cannot be traced to their statements. The report gives four examples of this, one of which relates to Professor Kibreab. It refers to the following conclusion in the FFM report,
“It is now
possible for evaders and deserters who have left Eritrea
illegally to return if
they pay the 2% tax and sign the apology letter at an Eritrean embassy. Kibreab
was aware of a few deserters from the national service who have visited
Eritrea
and safely left the country again.”
The report states that according to the documented conversation that the authors of the FFM had with Professor Kibreab, he followed this sentence with the following qualification: “These are invariably people who have been naturalised in their countries of asylum.” This qualification is not, according to UNHCR, included in the main text of the report on any of the three occasions that the statement is quoted.
There are three other examples of similar problems with the report which do not relate to Professor Kibreab.
7. A printout from
EIN summarising the UN human rights report on Eritrea
which was published on 8
June 2015. It is summarised as follows, “UN finds
Eritrea
responsible for
systematic, widespread and gross human rights violations, calls for
international protection for those fleeing”. The summary by EIN states,
amongst other things, that the FFM report followed a Fact-Finding Mission
undertaken due to a large increase in Eritrean asylum seekers in Denmark and that two Danish Immigration Service employees who were critical of the report
resigned in protest.
The Findings of the First-tier Tribunal
8. Judge Holmes
heard evidence from the appellant but he did not find him credible and he
rejected his evidence that he illegally left Eritrea
and that he was a
deserter. He made findings at [43] – [55] of the decision and it is necessary
for the purposes of this decision to replicate the following paragraphs:
“43. As set out above I have had
regard to the country guidance case of MO, and to the earlier cases of MA and GM
in assessing the weight to give to the evidence before me. I have also
considered the Country Information reports of March 2015, which rely (inter
alia) upon letters from the British Embassy in Asmara dated 1 April 2010,
and 11 October 2010, and, the Danish FFM report of December 2014 ‘Eritrea
–
drivers and root causes of emigration, national service and the possibility of
return.’ The Embassy letters were considered in MO, but plainly the Danish report is also based upon much
more recent information from a range of apparently reputable and reliable
sources, who might be expected to have detailed and first hand knowledge of the
information given to the authors of the Danish report.
44. I have considered the bundle of reports relied upon by the Appellant that offer criticisms of that Danish report from a wide range of authors. Much (although not all by any means) of that criticism is dependent upon Professor Kibreab’s own criticisms of the way the information he provided to the Danish FFM has been handled. The Appellant’s bundle does not include the statement published by Professor Kibreab on the internet of 25 March 2015 which offers his own criticisms of the Country Information reports of March 2015, but I am aware of its content. The Appellant’s bundle does not include the press release issued by the Danish authorities of 9 December 2014 detailing their chronology of their exchanges with Professor Kibreab, the occasions upon which he agreed notes of meetings and conversations held with him, and the occasions upon which he failed to respond to requests to do so, culminating in his email of 25 November 2014 to the Danish authorities congratulating them on a well written informative report, so that it was only on 28 November 2014 following its wider publication that Professor Kibreab sought to distance himself from that report. However that information is set out in section 1.3 of the report, and of course the quotations of Professor Kibreab’s evidence have not been redacted from the report, but merely struck through so that the reader may see in their proper context what they were. However much of what is now struck through, and withdrawn by Professor Kibreab appears to be a repetition of the evidence that he gave to the Upper Tribunal in MO [cf 24-39].
45. It seems to me
clear that there is a wide ranging dispute over the reliability of the Danish
FFM report of December 2014, and in turn over the Country Information reports
of March 2015. That dispute is centred upon the behaviour of Professor Kibreab
and the information he has provided. Professor Kibreab has for many years held
himself out to be an expert upon Eritrea
and what occurs in that country, and
to have been accepted as such by the Upper Tribunal. Put simply, if he was
accurately quoted in the Danish report as his own email of 28 November 2014
appears to accept he was, then he has undertaken a rather surprising and
complete change of heart following the international publication of that
report. That begs a number of questions about his reliability, and his current
stance towards the Danish report.”
9. The Judge
found that the applicant’s evidence was “inherently incredible” or “simply
inconsistent” with the evidence reviewed in MA and in MO. He went on to find that the appellant was
not a reliable witness in relation to any of the details of his account and
that he had created a fictitious account of his experiences and family
circumstances in Eritrea
.
10. The Judge
concluded at [49] that the appellant accepts that he is not a draft evader but
it does not follow that he is a deserter from national service or that he will
be perceived as one on return and the Judge concluded that he was not satisfied
that the appellant was a deserter. He went on to conclude at [51] that he was
not satisfied in the light of either MO
or the 2014 FFM report that there is a real risk that the appellant will be
regarded upon return as someone who left Eritrea
illegally, or as a deserter.
Error of Law
11. Judge Holmes applied MO but found that MST was not at risk on return. However, in assessing credibility he also considered the FFM report and the position of Professor Kibreab. The Judge’s conclusions about this evidence, namely that Professor Kibreab was unreliable and that MO was out of date in the light of the fresh evidence, informed his overall credibility assessment. Whilst it is not possible to determine the extent of influence this had on his assessment of credibility, it is clear that the Judge attached significant weight to the evidence relating to communication between DIS and Professor Kibreab and the unfavourable view he held about Professor Kibreab. Although the findings are framed in the alternative (see [51]), it cannot be discounted that had he taken a different view about the fresh evidence, he would have found the appellant credible.
12. The Judge did
not take into account all of the evidence in reaching conclusions about the FFM
report. Whilst the Judge properly concluded that not all the criticisms of the
report depend on Professor Kibreab, he did not adequately engage with the wider
evidence. There was before the Judge evidence from sources other than
Professor Kibreab that was, by any account, capable of undermining the FFM
report. There was a failure to properly engage with this evidence
(particularly the highly critical evidence from UNHCR and the newspaper article
of 10 December 2014 in which it is asserted that that DIS had stated that the
reaction to the report raises doubts about whether there are risks to people
returning to Eritrea
after illegally leaving the country and avoiding national
service). I am concerned that the Judge relied on a statement which was
published by Professor Kibreab on the internet on 25 March 2015, but this was
not produced by either party. I have not seen a copy of this statement. It is
not apparent what is contained in the statement and what weight the Judge attached
to it.
13. For the above reasons the FtT materially erred and I set aside the decision (in its entirety) to dismiss the appeal on asylum grounds.
14. This case will remain a Country Guidance case and is listed for four days on 25 April 2016. Following this the appellant’s individual appeal will be determined. The parties remain subject to directions issued by the Upper Tribunal. The appellant remains the subject of an anonymity direction.
Direction Regarding Anonymity – Rule 14 of the Tribunal Procedure (Upper Tribunal) Rules 2008
Unless and until a Tribunal or court directs otherwise, the appellant is granted anonymity. No report of these proceedings shall directly or indirectly identify him or any member of his family. This direction applies both to the appellant and to the respondent. Failure to comply with this direction could lead to contempt of court proceedings.
Signed Date
Upper Tribunal Judge McWilliam
Upper Tribunal
(Immigration and Asylum Chamber) Appeal Number: DA/00924/2014
THE IMMIGRATION ACTS
Heard at Field House |
Determination Promulgated |
On 22nd January 2015 |
|
|
………………………………… |
Before
UPPER TRIBUNAL JUDGE COKER
Between
AA
Appellant
And
SECRETARY OF STATE FOR THE HOME DEPARTMENT
Respondent
Representation:
For the Appellant: Ms A Benfield, counsel, instructed by Ziadies solicitors
For the Respondent: Mr M Shilliday, Senior Home Office Presenting Officer
© CROWN COPYRIGHT 2016
DECISION ON ERROR OF LAW AND DIRECTIONS
1. The appellant is an Eritrean national born on 6th September 1979 in Sudan. At age 3 months he moved to Saudi Arabia where the family lived with a residence permit. He arrived in the UK on 25th March 2003. He claimed asylum on 27th March 2003 and claimed his date of birth was 6th September 1986. His claim was refused and his appeals to the First-tier Tribunal and the Upper Tribunal were dismissed. He made an application for indefinite leave to remain on 15th September 2004. On 17th August 2007 he was convicted of sexual assault at Isleworth crown Court for an offence committed on 8th September 2006. On 21st September 2007 an order was made under s37 Mental Health Act 1983 authorising his detention at Three Bridges Unit. An order was made under s41 adding a restriction to the Hospital Order without time limit. On 21st February 2008 the appellant’s representatives made submissions on Article 3 and 8 grounds requesting that they be treated as a fresh claim enabling a right of appeal if refused. On 22nd July 2013 the respondent wrote asking for reasons why the appellant should not be deported and on 7th May 2014 a decision to deport was made pursuant to s3(5)(a) Immigration Act 1971.
2. The appellant was released from Three Bridges Unit to Chalkhill Facility on 10th January 2014 following a Mental Health Tribunal Hearing. Chalkhill Road is a supported accommodation placement with 24 hour supervision.
3. The appellant has been diagnosed with paranoid schizophrenia and is medicated on 800mg Quetiapine per day.
4. The First-tier Tribunal panel found:
a. the appellant has a history of dishonesty in relation to his asylum claim, including having initially claimed asylum as a minor, which he was not;
b. his parents continue to live in Saudi Arabia; he has 2 brothers and a sister there, a sister in Canada;
c. there is a house in Eritrea
formerly belonging to his grandparents and it is likely that his parents spend some time in
Eritrea
;
d. there is ample medical evidence, which is not doubted, that he suffers from chronic paranoid schizophrenia that is currently controlled by medication; he is monitored at Chalkhill round the clock but is able to go into the community during the day without supervision;
e. his Quetiapine medication is administered by the deputy Manager of Chalkhill; when he was attending to his medication himself he was not taking it correctly and his condition deteriorated; the appellant is aware that he has to take his medication regularly;
f. The panel referred to the findings of the Mental Health Tribunal that “there will clearly be risks in the community. He has no close relatives in the UK other than an uncle in Stockwell who does not visit him and he has no close friend in the community….He needs to continue with his medication to stay well”;
g. With residential supervision and regular medication he has settled into his current accommodation and can be trusted in the community without supervision during the day;
h. there is nothing adverse in his behaviour since the index offence in September 2006; he had previously received a reprimand for possessing an offensive weapon, cautions for theft of a bicycle and for possession of cannabis;
i. his criminal offence coupled with the need to supervise him and regularly administer medication in order to avoid a relapse with associated risks and the fact that he has no legal basis for being in the UK, make him a suitable case for deportation;
Error of Law
5. The appellant sought permission to appeal on four grounds:
1.
in
assessing Article 8 the panel placed inappropriate weight on the existence of
the extended family in Eritrea
, if they existed.
2. The findings by the panel as regards the possibility of forced conscription were not supported by the evidence and amounted to speculation
3. The panel failed to give adequate weight to the report of Dr Jaques and to the evidence of the respondent in determining the availability of adequate medication
4. The panel failed to apply the Joint Presidential Guidance No 2 of 2010 as regards vulnerable witness.
6. The appellant was granted permission to appeal the decision of the First-tier Tribunal in the following terms:
….The grounds of application argue that the panel erred in its assessment in relation to articles 3 and 8 of the human rights convention.
The grounds relating to article 3
are in essence that the panel failed to properly assess why the appellant would
be regarded as medical unfit and therefore exempted from conscription into the Eritrea
military.
Although I am satisfied that the
panel had in mind paragraph 11.05 of the 4 September COIS report (which is in
the respondent’s bundle), it is not clear that the panel considered it in the
context of paragraph 28.15, which indicates that mental health issue go
unrecognised. It is arguable that the panel’s conclusion, which is reached by
inferring that a mental health diagnosis in the UK would be sufficient to
establish that the appellant is medically unfit to serve in Eritrea
, is legally
defective because it is applying western norms into a wholly different cultural
and legal context.
Although I am less persuaded by the arguments relating to article 8, where the challenge is in effect to whether the panel properly carried out a balancing exercise, as I have found an arguable legal error, I will leave those issues open for the Upper Tribunal to decide.
7. Before me the appellant withdrew reliance on the fourth ground of appeal. The appeal before me was thus on Article 8, Article 3 (forced conscription) and article 3 (health).
8. Dr Jaques in his
report stated that the drugs available in Eritrea
(Risperidone, Fluphenazine,
Haloperidol and Chlorpromazine), information as to which had been received and
produced by the respondent in response to a Country of Information request, are
not appropriate options for the appellant’s treatment. The Deputy manager gave
evidence that if the appellant were subjected to enforced removal there would
be a rapid deterioration in his mental health and stability. She said that
within a couple of weeks there would be symptoms.
9. The panel found
76. There is some treatment
available in Eritrea
, though it is clearly inferior to the treatment the
appellant receives in the United Kingdom. It does not appear that the drug
Quetiapine is available in
Eritrea
. Risperidone may be available but is
unsuitable. ….it is for the individual to prove that medical treatment or care
will not be available to him in the receiving country. In this case we do not
have a complete picture of the drugs available in
Eritrea
. It has not been
shown to our satisfaction that there is no suitable drug available for the
appellant.
77. We do not consider that suicide is a real risk….
78. We do not find that the high threshold in Article 3 cases is met in this case.
79. It is common ground that the appellant is of conscription age and he might be eligible to be conscripted if he satisfies the age criteria and is ‘medically fit’. In this case the appellant is not medically fit, and the appellant would presumably have with him evidence of his treatment in the United Kingdom to establish that he is not fit. The argument that the appellant might be conscripted anyway because the authorities do not understand his illness or because the appellant might present as a healthy young man seems to us to be complete speculation. The fact is that the appellant is not fit to serve in the Armed Forces. We do not consider that conscription is a real risk in his case.
10. The OGN confirms
that Eritreans, who satisfy the age criteria and are medically fit, are subject
to conscription. The First-tier Tribunal determination is predicated upon
firstly the appellant being able to state what his medical condition is,
secondly that his statement will be accepted as a correct indication that he is
not medically fit, thirdly that the medical documents he produces from the UK
in English will be accepted as indicative that he is medically unfit and
fourthly it is a fact that he is not medically fit because he is not. The
medical evidence before the First-tier Tribunal panel included evidence that
whilst on medication and complicit he presents in such a way that he is able to
operate within the community without difficulty. He would not present as
medically unfit. The medical evidence was also that the drugs available in
Eritrea
were not appropriate for the appellant, that he was unable to self
medicate correctly, that a failure to medicate would lead to a rapid deterioration
(in the region of two weeks) which had consequences not only for his personal
presentation but also in terms of aberrant behaviour. The background country
material indicates a serious shortage of treatment available for mental health
problems. It was the Secretary of State’s evidence as to the drugs available;
to extrapolate from that evidence that the appellant had failed to prove there
were no other suitable drugs available was not based on a realistic premise.
The Secretary of State made an enquiry through her own respected channels,
disclosed that evidence and there was undisputed medical evidence that the
drugs were not appropriate for this appellant. It is difficult to understand on
what basis the First-tier Tribunal were able to find that the appellant’s
ability to describe his illness would be sufficient to enable him to be found
medically unfit. There is no background material or authoritative case law that
sets out how medical assessments are undertaken, whether information provided
in English is taken account of or what level of incapacity is deemed sufficient
to prevent conscription. The First-tier Tribunal did not engage with what would
happen at the airport on arrival: if questioned immediately on arrival he
would, because he would be on his supervised medication, present as medically
fit despite explaining he had serious mental health problems. If that resulted
in him being immediately conscripted, there has been no engagement with whether
it is possible to subsequently be examined for fitness and how that occurs.
There was no engagement with the consequences to the appellant if he were
conscripted, was unable to access adequate medication and his behaviour
deteriorated and what the consequences of that would be.
11. These are not, on the basis of the unchallenged medical evidence, speculative assumptions as to deterioration. The failure of the First-tier Tribunal to make findings on treatment on arrival, whether he would be detained and consequential treatment is an error of law. Whilst it may have been difficult to reach conclusions based on the evidence presented, there has been no proper engagement with the matrix of factors applicable to this appellant.
12. The First-tier Tribunal erred in law in its assessment of the Article 3 risk on forced conscription grounds.
13. In so far as the
Article 3 health findings are concerned the First-tier Tribunal erred in
finding that the appellant had failed to prove there was no appropriate
medication available. There was no assertion by the respondent that there were
other drugs available; the respondent relied upon the drug availability she had
ascertained was available. Those were not appropriate for the appellant. The
tribunal made findings that there was a family home in Eritrea
and it was possible that his parents spend time there and that there was some, albeit
inferior, mental health treatment available. Although referring to relevant
case law, the First-tier Tribunal did not engage with the specific facts for
this appellant namely that there was no evidence that there were any appropriate
drugs; that although there may be a family home there was no finding on the
quality or quantity of family care that may be available; what the consequences
would be for the appellant given the likely deterioration in health. Matters
relating to conscription would also be factors to be taken into account. Those
are all matters that require specific consideration in terms of the threshold.
It may be that the applicant does not meet the very high threshold required but
the First-tier Tribunal erred in law in failing to consider these issues.
14. In so far as Article 8 is concerned the grounds relied upon are in essence disagreements with the weight placed upon various elements of the appellant’s evidence. The decision reached by the First-tier Tribunal was well within the range of decisions open to it. There is no error of law in their Article 8 decision.
15. In conclusion therefore I am satisfied that the First-tier Tribunal erred in law in its decision on article 3 both in terms of the health issue and conscription. I set aside that decision to be remade.
16. On conclusion of the hearing before me on 22nd January 2014 I canvassed with the parties the future conduct if I were to find an error of law such that the decision is set aside. It was agreed that the resumed hearing would be limited to submissions only but that both parties were at liberty to file and serve such further evidence as they sought to rely upon.
Conclusions:
The making of the decision of the First-tier Tribunal did involve the making of an error on a point of law in so far as the decision on Article 3 is concerned.
I set aside the decision dismissing the appeal on Article 3 grounds; the decision on Article 8 and asylum grounds stands.
Consequential Directions
The resumed hearing will be listed for submissions only. Both parties have leave to file and serve such further evidence as they seek to rely upon; service to be no later than 10 days before the date of the resumed hearing. Both parties are directed to file skeleton arguments no later than 3 days before the resumed hearing.
Anonymity
The First-tier Tribunal did not make an order pursuant to rule 45(4)(i) of the Asylum and Immigration Tribunal (Procedure) Rules 2005. I was not asked to make such an order and am not aware of any reason why one should be made.
Date 9th March 2015
Upper Tribunal Judge Coker
EVIDENCE OF PROFESSOR KIBREAB (PK)
Report on AA, 23 September 2015
1.
The
first report prepared by PK dated 23 September 2015 was at the request of AA’s
representatives. In addition to a number of questions about country conditions
in Eritrea
the report addressed a number of points concerning medical issues.
2.
PK
described his initial involvement in the DFFM Report, his disassociation from
it and the subsequent criticisms he made in the public realm. With reference
to the critique he wrote immediately after the report was released, he
reiterated that he believed the conclusions of the report were what the DIS
wanted to establish from the outset. He deprecated the DFFM for wrongly
assuming that no empirical knowledge on Eritrea
could be generated without
visiting
Eritrea
. He found it unsatisfactory that apart from himself and Ato
Kebede the DFFM Report does not identify any of its sources. He said that the
interlocutors who were representatives of western countries “have clear vested
interests in terms of stemming the flow”. He said the DIS team had distorted
most of the information he provided.
3.
PK
said that despite the DFFM’s evident flaws the UK Home Office based most of its
guidance on this report. He cited the IAGCI critique of the DFFM Report and
its recommendation that the two March CIGs should no longer be used pending a
review. He referred to his publication on 25 March 2015 of a commentary ”Some
Reflections on the UK Home Office Country Information Guidance Eritrea
:
National (incl. Military Service and Illegal Exit), March 2015”. PK
acknowledges that the two September CIGs, whilst still referring to the DFFM,
draw on it in combination with other country reports and was more nuanced and
covers the 2015 UNCOI Report.
4. PK then addresses the direction of national service, taking issues with the Home Office position that it is generally between eighteen months and four years, the medical exception. In relation to the return of Eritreans, PK said the Eritrean authorities would know if some had exited illegally and ill-treatment may result.
Report by PK, 29 February 2016
5.
After
outlining his qualifications and experience, PK reproduces almost word for word
the contents of his Report for AA, mentioning his criticisms of the DFFM for
its “unsound methodology and unreliable sources” and distortion of information
he provided; stating that he never conveyed that draft evaders and deserters
are no longer routinely subjected to severe punishment; that he had a “dense
network of informants inside Eritrea
” built up over time. In very similar
terms to his report on AA, he addressed the various responses that had been
made to the DFFM Report and the UK Home Office’s heavy reliance on it for their
March 2015 CIGs, the subsequent critique by the IAGCI and Professor Campbell of
these documents, and his recognition that the September CIGs were more
nuanced. He reiterated that national service remains indefinite.
6. PK then turned to further instructions he had received asking him to comment further on other “wider” issues which the Upper Tribunal had directed to be addressed.
7.
As
regards illegal exit and its consequences he said he rejected the assumption
that those who left Eritrea
illegally no longer face any risks of persecution.
He considered this assumption to underpin the DFFM Report and the Home Office
CIGs including the September 2015 CIGs which in common with the March ones he
found “uncharacteristically of lower standard and devoid of evidence”. He
commended the Landinfo Report of March 2015 which had warned about the problem
experienced in obtaining information about
Eritrea
and the likelihood of
“round-tripping” or fake confirmations.
8. PK stated that the overwhelming majority of those who leave the country do so illegally because the issuance of exit visas is highly restricted. Having cited the categories of persons cited in the EASO Report of 2015, PK said that with the exception of children the categories were consistent with those he had identified to the Tribunal in MA in 2007. His own findings showed that the Eritrean government denies exit visas to children between five and seven. He considered that the category of former freedom fighters and their family members was too broad as it was really confined to those enjoying connections to the President and his inner circle. The same was true of authority representatives.
9.
In
his conclusions PK added that whilst category (iv) of MO (a person
declared by an official committee to be exempt from national service on medical
grounds) remained the same, the government had introduced stricter controls for
MO category (v) (those seeking medical treatment abroad). Owing to the
need for specialised skills required by government departments, there are
people who can get visas for skilled work, but the process of vetting based on
loyalty has become more severe. The number that leave for the purposes of
further training has diminished considerably because of the numbers leaving
unlawfully and those who leave lawfully but fail to return. The repressive
nature of the regime in Eritrea
made it impossible to document what happened to
failed asylum seekers, but it was “safer to assume” that deserters would be
detained incommunicado indefinitely. The only evidence-based case he knew of
concerned the posting on the Eritrean website assena.com on 16 February 2016 of
the case of Berhane Embaye who had fled to South Sudan whilst serving in the army
and on being “cajoled” by the government to return, disappeared soon after his
arrival. This action, PK considered was “typical of the treatment the Eritrean
authorities accord to those whom they suspect of disloyalty”. The Eritrean government’s
claim that only those who did not commit an offence are exonerated mean that
only those who left legally after being demobilised are unlikely to face
punishment. Payment of the diaspora tax does not immunise those who exited
illegally from punishment. The reason why the Eritrean government introduced
the repentance letter was to discourage exiles from joining opposition groups
in the diaspora: it has no other significance. It amounts to a willingness to
accept punishment.
10. As regards those Eritreans who were returning to visit, they were naturalised and the Eritrean authorities did not detain them except in exceptional cases.
11. As regards exemptions, he considered there had been little or no change since MA and the numbers were likely to be insignificantly small.
12.
PK
said it was possible for those who left Eritrea
illegally to obtain an Eritrean
passport, although many were reluctant to do so.
13.
There
was no evidence, PK said, that the Eritrean government has a more “relaxed”
attitude towards those returning to Eritrea
who left illegally. His statement
to the DFFM did not state this.
14.
Those
who left Eritrea
illegally but return when they are below draft age will be
required to do national service but are unlikely to face other consequences.
Those who return after reaching draft age are also unlikely to face other
consequences unless their reason for departure was to avoid conscription.
15. In the main body of his report PK said he considered that the Eritrean authorities still treated the benchmark age for imposing exit visa restrictions as eight years old, but in his conclusions he said the age was now reduced to five years.
16. As regards adults, PK considered that there was no longer a 54 year limit for men and 47 age limit for women. It was now 70 for men and 60 for women following the introduction of the people’s militia, although it was very unlikely that women over the age of 47 would be required to serve in the national service. Both men and women in the people’s militia can apply for permission to leave the country through their unit commanders.
17. As regards the question whether the requirement to undertake national service put the person at risk of persecution or serious harm, PK agreed with the characterisation by the ILO that the Eritrean national service regime involved forced labour. The hundreds of thousands of conscripts are seldom involved in military-related activities, most being involved in manual labour on construction sites, agricultural farms, housing projects belonging to the government and the PFDG as well as senior military officers. Many also work in the civilian sector of the administration. A person subjected to forced labour against his/her will under the menace of severe punishment is undoubtedly suffering persecutory treatment.
18. PK said conditions in prisons and detention sites were severe everywhere, most taking place in shipping containers and/or underground dungeons. Common problems mentioned by former detainees are beatings, overcrowding, lack of sanitary facilities and shortage of food, water and ventilation.
19. PK considered that the open sources and data gathered from former conscripts indicate that desertion is severely punished by the government, there are no rules applied consistently. Given the vindictive nature of the Eritrean authorities they are likely to punish re-conscripted deserters and draft evaders severely.
20. PK gave answers about exemptions from being conscripted for former liberation fighters, those declared unfit. He was adamant there had been no demobilisation for those within the age of conscription. As regards the people’s militia, it runs parallel to the national service and all citizens between 18 and 70 are eligible to join. Members are required to receive military training and carry weapons and are required to do unpaid manual work in development projects of different kinds.
21. Although the government’s plan is to recruit throughout the country, so far only some regions such as the Central, Southern and Anseba are affected. Those who refuse to respond to the call for service can face serious consequences such as loss of ration cards and imprisonment. There are no exemption categories. The consequences for those fleeing the country to avoid the people’s militia and then returning are likely to be the same, “but I am speculating”.
22. Some of PK’s conclusions have already been noted. His two principal conclusions are that:
“ There is no evidence to suggest that the Eritrean authorities’ hostility to those who desert from or evade the Eritrean NS [national service] and subsequently flee the country has diminished”
and
“ In light of the
restrictive categories of those allowed to leave Eritrea
lawfully, the great
majority will be perceived as having left illegally. Save for very limited
exceptions, returnees will face serious harm. This will not be avoided by the
payment of a 2% tax and/or letter of apology. Rhetoric notwithstanding,
nothing has changed on the ground.”
23. The professor’s report was subsequently modified in minor respects on 4 April 2016.
Reflections on Home Office FFM, 21 April 2016
24. PK has also provided a critique of the Home Office FFM based on visits in February 2016. He takes issue with the presence of a representative of the MoFA in most of the interviews and the presence of an interpreter provided by the MOFA during the interview of returnees interviewed in Tesseney. He considered that a good number of the interviewees had to be viewed as pro-government sources unlikely to be objective. The fact that these interviewees all seemed “to sing from the same hymn sheet” reinforced that view. He considered the quality of the three anonymous interviewees was “indisputably far superior to any other data the FFM gathered during the mission”. PK then comments on the contents of a number of interviews, notably agreeing with the British Ambassador on some points but not on others. He was particularly disparaging about the evidence given by the head of the government’s political officer, Yemane Gebreab.
25.
In
the course of dealing with some of the UKFFM interviews, PK offers further
evidence of his own. For example PK said it has been the policy of the
Eritrean government for some time that, if a person who lives abroad and is
within the age of conscription returns to Eritrea
and stay for a year or more,
he will be required to do national service. He said he knew of several people
affected by this policy who routinely leave the country for short periods and
return with stamps on their passports of foreign countries to show they have
been away. “This may also be seen in the exit visa issued by the Eritrean
authorities when they leave the country”.
26.
In
reaction to the UNMS interview, PK considers that those going back to Eritrea
on holiday “is a small proportion which is close to the regime”.
27. PK concludes that in the absence of verified data and in the light of the government’s prevailing dismal human rights record and the dearth of political and economic reform, it is safe to assume that those who fled illegally and are forcibly deported would face persecutory treatment.
28. It was PK’s view that the data in the notes of the FFM indisputably indicates that the national service is devoid of uniform rules, which is also consistent with the findings of his own studies over many years. He was critical, however, of the virtual silence in the FFM notes about the prevalence or absence of corruption. He found it difficult to accept that those working for embassies and mining companies were 100 per cent demobilised. Round-ups and the shoot to kill policy continue.
PK’s answers to written questions submitted on behalf of the Secretary of State
29. In reply to written questions from the respondent seeking clarification of aspects of his report, PK sent his answers in April 2016. He noted and regretted that his initial report did not conform in all respects with the Practice Direction of the Immigration and Asylum Chamber of the First-tier and Upper Tribunal, 13 November 2014. He sought to clarify various aspects of the use of sources and informants who included refugees and asylum seekers, the latter whose information was likely to be more up-to-date. PK accepted that when the DIS team had asked him for feedback on the draft DFFM they had sent him, he had emailed “Thank you for the well-written report” without even opening the email attachment. He was under great pressure running three MSc programmes at the time. He believed that the head of the DFFM had then agreed to give him time to send corrections to what he had by then learnt were distortions. However, Mr Glynstrup then emailed back saying they were going public immediately.
30. PK said his criticisms of the DFFM Report were of the main body of the report but the problem remained with the interview material namely that Messrs Olsen and Olesen had cast doubt on whether the underlying interviews were themselves reliable, given the use of leading questions and Mr Glynstrup’s fixation with achieving a specific result.
PK: Oral Evidence
31. Much of PK’s oral evidence covered the same grounds as his written evidence. He explained that he had prepared his report for AI before he was aware it would be used for the country guidance case; he did not learn he had been jointly instructed by all three appellants until 23 January 2016. In April he made revisions in his February 2016 Report so as to add a ‘Conclusions’ section. He accepted that this had led to him omitting paragraph 1.5; he did not know how this had happened. He accepted he should have made an effort in his February Report to cite the correct Tribunal Practice Directions.
32. PK said that when concluding his research in 2013 he assembled his 190 respondents by using the snowball or “chain referral” technique. It was indicative rather than conclusive but one could rely on it if properly carried out. The limitation is that you cannot get as diverse a group of respondents but he had a built-in method for correcting this by identifying multiple sources (using different ethnic groups etc). He accepted that if not used with care this technique could result in “roundtripping”.
33.
Asked
about the extent that his research on Eritrea
was up to date, PK said he had
conducted about 20 interviews since 2012 and a further four interviews with
Eritrean sources since 3 March 2016. These 24 came from different backgrounds;
they were all people who had fled national service; all were under 30, some
women, some Christians; some who had fled as recently as end of 2015 and two
who had fled in 2016. He gave details of two examples. As regards the three
returnees he had referred to in his DFFM interview he said he knew one
personally who had been in national service for more than five years and was a
very active supporter of the government. He had met him through his cousin in
2013. He went back in 2013. He had good connections with the government and
was an active participant in the ruling party. He understood he had got
clearance from the Eritrean embassy. He did not discuss with him whether he
paid the diaspora tax. The second person he had found out about through a
friend and he too was very active in the ruling party. He did not know him
personally. He had no discussions about the diaspora tax. The third person was
related to a friend of his and he had been granted asylum in the UK He was
actively involved with the Eritrean embassy. His uncle was a prominent freedom
fighter.
34.
Mr
Rawat asked was it unusual to have conversations with Eritrean people about the
topic of Eritreans going back to Eritrea
. PK said it was one of the most
contested issues people talked about. Most objected that it undermined the opposition
parties, as it would provoke the reaction, “If they are safe there, what are
they doing here?” It was seen as undermining the opportunity of others to be
granted status.
35. PK said he had read the new UNCOI Report and had noted its reliance on fresh evidence.
36.
Mr
Rawat turned to the subject of the DFFM Report: PK said he was interviewed by
the DIS researchers in September 2014 – Mr Olsen and Olesen came to see him
twice. They sent him their draft transcript and he approved amendments on 14
November 2014. When he was approached by Olsen and Olesen he was happy
to cooperate as he considered that their concern was to investigate the human
rights situation in Eritrea
and they were very honourable. Mr Rawat asked him
why he had written that he had felt “used and betrayed” when he himself had
approved the note of his interview.
37. He said that his statement about a relaxation in the policy of the Eritrean government towards returns had been taken out of context and cited three times in the 20 page main body of the report (see para [175] above). He had not had time to check through any of the report because he had heavy academic commitments at the time. He had told the head of the mission that he could not comment quickly yet they had gone ahead. At that time he did not know that Olesen had resigned and so was unaware he would not be given more time. He had made a mistake. When he had emailed the head of mission on 25 November 2014 that the report was “well-written and informed” he had not read it. He conceded he should not have done that. The DIS head of the mission had taken advantage of him. But he had a clear conscience and all COI professionals had supported him. Mr Rawat asked whether he believed he could give independent evidence when he was so intimately connected to the event surrounding the DFFM and when he described himself as a victim of unethical conduct and someone who had been “betrayed”. PK said his ethical standards had been untarnished throughout. He believed he had been objective. He had never been emotionally involved. The report ended up being discredited and that was nothing to do with his intervention.
38.
PK
said that he accepted that fact-finding missions were in themselves a valid
type of exercise in the context of Eritrea
so long as they did not seek to
synthesise and analyse the date. He applauded them (the FFMs) for trying
different sources.
39. When asked whether he thought that the DFFM was completely flawed or whether it was possible to differentiate between the first 20 odd pages and the sections which simply gave the texts of the interview transcripts, PK said that from the statements of Olsen and Olesen the contents of some of these transcripts may have been affected by the methods used by their boss. The section of the report which describes the sources and what they said is contaminated and the quality of the data effected. There is generally no harm in FFMs. They are not a complete waste of time, but there are caveats about anonymity and sources. The Danish FFM is exceptional and he would not put the UKFFM in the same category.
40. He agreed that the Martin Plaut article was not an independent “analysis” (as he PK had described it in his report) since it was cut and pasted from the HRW press release. He had used other sources and had not relied just on this source.
41. Mr Rawat questioned why PK had not mentioned in his ’Reflections’ document the Home Office response to the IAGCI Report criticising the March and September CIGs. PK said he knew of its existence but saw no reason to cite it. Mr Rawat asked PK if he did not consider that his duty as an expert was to mention both sides to any issue. He said he agreed with Professor Campbell. He read a lot when he wrote a report and only referred to other reports when he considered it relevant. He did not change his mind after he read the Home Office response. He did not consider it necessary to reflect on the Home Office view regarding use of anonymous sources.
42. Asked about the methodology used by the UNCOI in its first Report of 2015, PK said he did not think the degree of anonymity was excessive, but agreed that as a result very little was known about them. For Eritreans there was always anxiety that the government surveillance would make them identifiable and their families could face repercussions. He did not know how the informants were selected.
43.
PK
was asked a number of questions about the methodology of the AI ‘Just Deserters
Report’. He could not say if the “range of sources” this report drew on were
obtained through use of the snowballing technique. He accepted that it
appeared that the ‘Just Deserters Report’ had two persons (Filmon and Yonas)
saying identical things, but it was possible it was two persons who fled
together or had identical experiences or who had chosen to say something
jointly or who had rehearsed the same story. If it had been anyone else than AI
he would have been concerned there had been error. It was incompatible with
his knowledge of AI that they would make such an error. He accepted there were
two other passages, virtually the same from two persons Elan and Danait. He
accepted that if AI had annexed transcripts of these interviews it would have
been possible to check. He did not think this cast doubts on other parts of
the ‘Just Deserters Report’. He would give the benefit of the doubt. He knew Eritrea
and what was described here matched his knowledge of the country. Generally the
report was accurate about the country. He did not need to rely for his own
assessment on the ‘Just Deserters Report’ in any event. He counter-checked
with other sources.
44. PK said he accepted that AI’s decision to rely on the evidence from asylum seekers meant that questions had to be asked about any possible incentive to exaggerate, but that did not mean that their evidence could not be relied on. He preferred to use people with status. AI do not have to apply a scientific method at all times. If Mr Rawat wished to challenge the AI evidence he should ask them questions.
45.
Asked
why he had not mentioned the changes in age for lawful exit in his February Report
(these were added in the April revision), PK said he probably made the changes
as a result of the research he did. Asked why the only source he cited for
this change was a book published in 2013 (The African Garrison State) which was
not in any event sourced, PK then said that it was not the only source but
there were internet sources and that he had relatives in Eritrea
who were over
70 carrying guns. He referred to oral communications and common knowledge
which he accepted that he had taken for granted and accepted that this was not
“right.” It was put to PK that not only was the source he used inadequate, but he
had had failed to mention the source in his February 2016 Report and he was
unable to give an explanation about this.
46.
PK
said he had other sources, e.g. the BBC. He agreed he should have given a more
recent source. He had had oral communications with several persons in Eritrea
that the age limit for women was now 60 years because of the people’s militia. He
had friends and relatives affected by this. It was common knowledge. The fact
that a person was not actually engaged in the people’s militia did not exempt
people. It would only be a matter of time before they would be asked to carry a
gun. He had heard Eritreans talking about the possibility of women under 60
being granted an exit visa, but he was not aware of anyone himself. He was not
aware that Landinfo had said that one factor affecting whether men and women
aged under 70 and 60 respectively could get an exit visa was whether they had
done mandatory weapons training. In any event, being able to apply and being
able to get were two different things. He would not rule out someone being able
to obtain a visa because there is a lot of corruption and it is arbitrary.
47.
Asked
why if he considered the USSD Report should be seen as relevant in relation to
the age limit for children (five) he did not agree with its April 2016
assessment that women over 30 could get exit visas, PK said that was not his
information. He had recently had to help a female relative in Eritrea
over 40 years old and she had been told she was not eligible because she was
potentially liable for the people’s militia. He accepted that this amounted to
evidence of one person and that he was generalising. He did not mention this
communication in his report. Although the people’s militia was part-time you
had to be available. His aunt is in the people’s militia and is not engaged on
a full-time basis but has to be available whenever there is a specific task and
can be called up at any time. He said he had never found it necessary to
investigate the matter. To say the age limit for women was now 60 was not a
“guestimate”. He did not agree with Mr Bozzini’s 2012 analysis that women when
they reach 27 can regularise and demobilise. People with connections might be
able to achieve this, but it was not policy. He did not agree with the
respondent’s case that the upper age for women doing compulsory national
service had in fact decreased to age 30.
48.
Asked
further about his treatment of the lower age limit for children, PK accepted he
had said eight years in one place, between five and seven in another and was
now saying five years. “We are not dealing with exact figures in Eritrea
”. The USSD Report used the term “generally”. He considered it was appropriate to
say between five to eight or eight. Leaving aside the USSD Report which did
not cite a source or sources, PK agreed the only source for the figure of five
was the immigration officials interviewed by the UKFFM in February 2016.
49.
Mr
Rawat asked PK what his sources were for saying in his April ‘Conclusions’
section that the Eritrean government had adopted a stricter approach to exit
visas. PK said this is what he had heard from friends and relatives. His
information was just anecdotal. He did not accept that this position was
inconsistent with his evidence about Eritreans travelling to Sudan for medical treatment. An exit visa is not needed to travel to Sudan for those ineligible for national
service or the people’s militia. He did not agree with the evidence of the
immigration officer interviewed by the UK FFM about this. People needing
medical treatment can travel to Sudan there being many buses making the
journey. “Were the people on these buses confined to those over 70 or under
[five]?” asked Mr Rawat; PK said the visa regime was stricter for those
affected by national service or people’s militia. He expanded on this when
questioned by Ms Dubinsky. He said that to go to Uganda or Kenya a person needed an exit visa, but for Sudan a travel permit would suffice. The criteria were
the same for those within the national service age ranges, but outside of
these, e.g. for elderly people it was much easier. A person who goes to Sudan without a travel permit would be regarded as having left illegally. He thought a
significant part of the movement between Eritrea
and Sudan comprised older
people including former professionals who had established their own businesses
and had left for South Sudan but still had family ties back in
Eritrea
, they could get documents.
50. The bench asked PK to expand on his evidence about travel to and from Sudan in light of his earlier evidence that there was presently a lot of traffic between the two countries. Were those on the buses just people under five and over 70, he was asked. He said that usually the people would be older, they would be holding certificates from the war of independence and notes from a doctor; but all people would need travel permits. The border is very porous and people can easily cross and that is why so many people leave illegally. Asked if there would be a lot of checking at the border exit points to ensure those going to Sudan were not evading or deserting national service, he said that legal exits could only take place in particular places. Check points on the border are very limited.
51. Asked whether he had any view of the accuracy of the Eritrean government estimate that between 60,000 and 80,000 exit visas were issued per year, he said the range given did not suggest these figures were based on fact but it was not within his knowledge.
52. PK said the suggestion in some diplomatic sources that returning Eritreans used Eritrean passports was misplaced.
53.
Asked
about his citation that fewer students could now get exit visas, PK said this
was his observation on the consequences of more Eritreans leaving the country
and the introduction of people’s militia service. In 2014 the government had
realised the scale of the numbers leaving was untenable. One of his sources was
awate.com, which is run by an opposition group and he considered credible; but
there were many other sources, although none of them specify anything about
students. He agreed that what was reported in awate.com was that there had been
a reduction in the size of the militia and the army and that there was in fact
no reference to students or scholarships, but that he had deduced from this
that awards of scholarships had significantly diminished because Eritrea
relied on conscripts. There was evidence of Eritrean students studying in the Gulf
states, China, the Far East and other African states, but very few students
return, so those granted were really only supporters of the government and its
party. But as regards applicants applying to study in Europe between 2011-2015,
he had no source of evidence as to whether numbers were going up or down; he
just relied on common knowledge.
54. As regards his claim that the number of authority figures able to get exit visas had shrunk, he said he based this on phone calls he had had with such people, but he could not quantify.
55. In reply to questions about demobilisation PK said he used this term to denote complete demobilisation. Relocation to national civilian service was not demobilisation. The only basis for demobilisation was ill-health – unless you factor in payment of bribes. He did not accept there was a process for requesting demobilisation. If there was such a process (as suggested by the UK Ambassador) it was a corrupted process. He would know from his sources if there was such a process. There are no rules that regulate the process. He did not accept that there was de facto demobilisation for married or pregnant women. They were not called up to active military service, but they would be called into civilian national service, including women over 50. However some women, particularly those who are educated, are still called up to work in offices. Women were not formally demobilised but may not have an assignment; the issue only became relevant if she wanted to travel. You may see women prohibited; there was a lot of arbitrariness. In the majority of cases women could not get a discharge certificate but a minority may get travel permits and these were not generic documents but were specific to the time they were applied for.
56. PK said he considered that the 2015 UNCOI evidence about ad hoc exemption and the existence of a “certificate of completion of [national service]” was consistent with his own analysis. Some may be able to get these documents. Exemption certificates would not mean persons got permission to exit, different rules applied. He did not think that centralised records existed of those de facto allowed to leave national service. That was why there were round-ups. Relatives of even pregnant women had to go to commanders to get permission. If a woman below 47 were returned forcibly she was most likely to be treated as a deserter and would not be able to prove exemption on the basis of de facto demobilisation. The punishment could not be predicted. There was no regularity
57.
In
cross examination Mr Rawat asked PK about the medical illness exemption from de
facto demobilisation. It was easier, he said, for people with visible
detectable illnesses. One of the real problems was that the processing of
people to establish whether they were unfit took time and ill-treatment could
happen meanwhile. His research in Eritrea
in September 2002 had included
conversations with a nurse. Since that research his knowledge of how the
Eritrean authorities dealt with mental illness came from his own life
experiences. (PK said later on that he had not retained notes from his
interview in 2002, he relied on memory and had sought to paraphrase. He did not
always cross-check). He did not accept the Minister of Health account to the
UKFFM that there was a qualified psychiatrist in St Mary’s hospital; he was
probably a G.P.
58.
Mr
Rawat asked about Eritreans working in the embassies in Asmara. He agreed that
the position was that these embassies could not employ those who had not
completed their national service/being demobilised, but based on his intimate
knowledge the government liked to insert their supporters in the embassies, so
the embassies would not know “the state wants the President everywhere”. Getting
documents in Eritrea
saying one was demobilised was easy enough, but it did not
mean one was actually demobilised. He could not defend this claim in a court
of law, but it was his understanding from discussions with fellow scholars and
friends. Perhaps in regards to international organisations the Eritrean government
would not be so concerned to have supporters inserted. He did not think any
credence could be attached to the UKFFM respondent he spoke to about being
seconded to Nevsun. He accepted that at the Bisha Mines employees had been
demobilised, but this was not true of the work done by subcontractors, over
whom Bisha Mines have no control certainly outside the main sites, who were
often companies run by leading members of the ruling party. This is common
knowledge. Segan is a sub-contractor and belongs to the ruling party and given
how secretive this company is he did not accept the claim they did not use
conscripts. He did not accept the findings of a Human Resources audit that in
his view had been commissioned by Nevsun (Human Rights Impact
Assessment of the Bisha Mine in
Eritrea
(2015 Audit) - Nevsun Resources Ltd, 5
August 2015) and he had not read it.
59. PK said he did not accept UKFFM materials that suggested it was possible for families with only one breadwinner to be demobilised. He had family members who were sole breadwinners, one of whom had been in national service for 20 years, another for 16-17 years. If there were a sole breadwinner rule, most Eritreans would be demobilised.
60.
As
regards the extent of the private sector in Eritrea
, PK said it was banned in
2006, and is now confined to local level small family businesses, but ability
to work in such businesses depended on the whim of the local commander, it was
not common. PK said he accepted that it was possible for foreigners like the
British Ambassador to walk around Asmara and even to travel outside with
permission and to observe events and to speak with people involved in
business. But the British Ambassador did not have a comparable ‘dense network’
of sources. He initially asserted that the likelihood of Eritreans divulging
information to foreigners was zero, but he accepted in oral evidence that this
was an overstatement and that he should have been more careful.
61. PK confirmed his position which was that the regime does not punish people who leave illegally, but those punished are deserters or evaders. Illegal exiters are not at risk unless perceived to be in this group.
62.
PK
was taken to his “Reflections” document where he had criticised for having a
“cavalier attitude” those who said that the “shoot to kill” policy was no
longer in existence or was applied less than previously. The sources cited in
support were not reliable. He believed reports about its continued existence
were the tip of the iceberg. He did not blame those who asserted this using
normal standards of evidence assessment, but Eritrea
is different. The UK Ambassador may honestly say this, but it was not a reality. Asked if he understood that
he was alleging in effect that the UK Ambassador was ignoring evidence, he
agreed he could not disprove what had been said, by “cavalier” he meant about
general violations of human rights. He could have put it in better language.
To say “single, isolated incident” with reference to an incident resulting in
the death of ten people (the shooting of conscripts in Asmara on 3 April 2016),
was appalling. It was not isolated; it was a pattern. At least the Ambassador
was only entitled to say there was a decline in reported incidents. Very few
people know what is going on. He accepted he himself was not in a position to
verify the 2014 incidents.
63.
When
questioned by the bench he said that perhaps “he had taken it too far” in
suggesting western representatives could not be objective, but he stated that
it is common knowledge that there are no political and civil rights in Eritrea
and that this did not mean that people have no right to claim asylum. He was asked
whether it is possible for western diplomats to be objective and he commented
on the rise in asylum seekers from
Eritrea
and that this has become a major
topic of concern, but that he would not dare to homogenise the diplomatic
community. There may be others that think they cannot afford to accommodate all
these people, but he conceded that he was speculating, using (he said) common
sense and intuition. His evidence is not that those compiling the CIGs are
incapable of being objective and having given talks to them in various
different location he is familiar with the preoccupation of these people. He
was asked whether the main interest of these people he describes is of stemming
the flow of immigration and he stated that what he said was an observation.
64. He stated that there may be those that think that the West cannot afford to accommodate Eritreans, but he conceded that he was “speculating and using common sense”.
65.
He
was not suggesting that those who compile CIGs are incapable of being
objective, but when he looked at the Home Office CIGs he considered the only
explanation for the dramatic change was a concern about using numbers of
Eritrean asylum seekers coming to Europe. Governments in the West are
determined to stem the flow of migrants and turn a blind eye to what is really
going on in Eritrea
.
66.
Asked
why he had relied on the Edmund Blair article of 25 February 2016 to justify
his view that recruitment to national service had become stricter, PK said even
though Blair had not referred to this, it was obvious from the article in which
it is asserted that the Eritrean government will not stop recruiting young
people into national service for lengthy periods. The Eritrean government
continues to have the rationale for being tough on national service, namely the
threat from bigger neighbours. PK confirmed when cross examined by Ms Dubinsky
that Eritrea
still considered itself to be in a state of emergency. PK said it
was the first time since 1998 that an Eritrean minister had said recently that
there was no threat, but the ‘no war, no peace’ policy was still in place.
67.
PK
was asked questions about his treatment of reports about the 3 April 2016. He
had phoned a contact in Eritrea
who had confirmed the incident. PK had
described the victims as draft dodgers, but he accepted that none of the
reports referred to “draft dodgers” and he apologised for doing so. It was
suggested to him by Mr Rawat that there had been an element of planning by the
conscripts because friends had blocked the convoy, but PK did not comment on
this.
68. Mr Rawat asked PK about his evidence that families of those who left illegally were no longer being asked to pay fines, or less so. It was less widespread. The government is unpredictable. However he gave examples of family members of others who have been arrested and detained.
69. PK gave extensive evidence in cross examination and following questions by the bench about the paragraph in the DFFM. We have referred to this when assessing the DFFM evidence (see [175] above) but it is necessary to repeat the text here;
“in the past two to three years the
government’s attitude towards national service seems to be more relaxed. It is
now possible for national service evaders and deserters who have left Eritrea
illegally to return to their country. They must go to an embassy and sign a
repentance letter in which they accept any penalty for the offence committed.
In addition they must pay the two per cent diaspora tax. Finally, they are
obliged to participate in festivals. In spite of this softer approach many
evaders and deserters still do not dare to return to
Eritrea
. Individual
circumstances play a role as well. Persons who did not participate in
oppositional political activities abroad and people who are connected by family
bonds or in other ways with government officials or members of the ruling party
would be more inclined to return to
Eritrea
on visits. Gaim Kibreab [PK] was
aware of a few deserters who have visited
Eritrea
and safely left the country
again. These are invariably people who have been naturalized in their countries
of asylum.”
70.
PK
was questioned extensively about this paragraph. He stated that the diaspora
comprised two groups; one that came to the UK during the war of independence
(they did not leave illegally and are not perceived as evaders/deserters) which
he described as the large majority and this is the only group who can safely
return (he later said that this large majority did not include evaders and
deserters who had fled Eritrea
illegally having initially said it did). He
confirmed that they did not have to pay the tax or sign the letter. They have
been naturalised by another state. The second group; the minority are the
others who cannot safely return, having fled illegally. He said that he did not
rule out those having fled illegally and having been naturalised returning (he
knew of three people in this category), but they would have to have links to
the government and anyone in this second group who was not well connected and
who did not support the regime would not return as they would not be safe. It
was this group that had to sign the letter and pay the two per cent tax.
71. He was not referring to the large majority in that paragraph.
72.
Those
who are able to return safely are those who have been naturalised, but those
who illegally left Eritrea
would be reluctant to return and would not in any
event approach an embassy. He did not accept the evidence of the UKFFM that
deserter evaders can safely return. He reiterated that he had never heard of
anyone returning for a holiday without having a foreign passport. He based
that on his own research.
73.
In
respect of “some people” he referred to in the paragraph who had visited Eritrea
and left safely, he knew of only three people in this group.
74.
PK
stated that the diaspora two per cent tax could be paid by those who left Eritrea
illegally, but the reality is only those who had regularised their position would
pay it. The payment of the tax and signing of the letter did not mean persons
would be able to return. It does not provide any immunity from ill-treatment.
The Eritrean government would look at a person’s activities and whether they
were connected with opposition groups and the like. PK re-emphasised that the
repentance letter required persons to admit to a crime, which reflected the
vindictive and arbitrary nature of the Eritrean state.
75.
PK
was asked if the signing of the regret letter and payment of the tax was
independent of gaining entry to Eritrea
. He said it was; these acts were not a
requirement for returning Eritreans to re-enter. He did not think that the
great majority of those returning for holidays were required to pay anything
and the same would be true of their children. However, Mr Knafler asked whether
Eritreans who left
Eritrea
before independence have to pay the diaspora tax he
said yes and that he had a friend who needed an ID card and she was forced to
pay the tax.
76. He was asked by the bench about any risk that members of the diaspora would have to confront and he said that if they stay longer than a year they would be required to do national service. He was asked whether people who left before the war of independence would have to pay the two per cent tax and he stated that they are forced to when getting an ID card.
77.
Dr
Bozzini was wrong in February 2012 to say many pay the tax; it is a small
fraction. Dr Bozzini was right, however, that this tax bought you access to
consular services, such as being able to buy land. Payment of the tax is not a
prerequisite to being allowed to enter. The diaspora tax gives you access to
services inside Eritrea
but you do not need to pay it in order to send
remittances, although the Eritrean government has tried unsuccessfully to block
informal channels of payment. The ICG report was wrong about this. He knew this
through his own family members and their experiences. In re-examination PK said
he had seen the UNCOI Annex VII letter of apology (regret) document before. He
did not know its source but he had seen it a long time ago. (Source: the 2015
UNCOI Report (Advance version of the Report of the
detailed findings of the Commission of Inquiry on Human Rights in
Eritrea
, A/HRC/29/CRP.1, United Nations General Assembly Human Rights Council), 5 June 2015).
78. PK was asked about the incident cited at [436] of the 2016 UNCOI Report which suggested that payment of the tax prevented ill-treatment. He considered this reference inconclusive. PK said that because such a person does not face adverse treatment may not be because he has paid the tax therefore implying that there may be other reasons. He implied that there was insufficient detail about the circumstances.
79.
He
confirmed his position in his DFFM interview that those who left illegally
could obtain a passport with reference to a facility introduced by the Eritrean
government in 2001-2002 as an exercise in reducing the appeal to such persons
of Ethiopian citizenship. However, it was not his evidence that those who left
Eritrea
unlawfully could, in 2014, approach an embassy and get a passport by
signing a repentance letter and paying the tax and that this would ensure safe
return. Unless confident that the government would protect you would not go
down this route.
80.
An
Eritrean ID card doubles as a visa. As regards Eritreans obtaining Eritrean
passports, this could be done abroad by those who needed them for onward travel
but the Eritrea
government could refuse applications. Legally speaking a person
returning to
Eritrea
on holiday in possession of a foreign passport could be
required to do national service in the same way as the resident population but
in practice the Eritrean government treated dual nationals differently.
81.
With
specific reference to the paragraph (quoted in full at [69] of this Appendix) and
the government attitude having relaxed, he confirmed that this applied to the
government’s attitude to supporters and members of the party, people with
connections in the context of those who had left illegally. As an example he
cited the three deserters (see above). However, he referred to the paragraph
and was emphatic that this applied “invariably to people who have been
naturalised in their countries of origin”. The caveat at the end of this
passage was important; those in this category were invariably those who had
naturalised. Despite what he said at 6.1 of his Report of 23 September 2015,
about there being “no evidence whatsoever” that the Eritrean government has now
a relaxed attitude towards those returning to Eritrea
who left illegally,
there was some evidence, but only for those who were well-connected.
82. He also further clarified that he had meant that only those who were members, supporters of the government or connected with the party could sign the letter and pay the tax. The great majority category that he had made reference to earlier are not required to sign anything
83.
PK
was asked by the bench to give more detail regarding the three persons who he
knew had returned to Eritrea
and whom he had in mind in his interview for the
DFFM Report. The men had been naturalised as British citizens having been
granted asylum. He said that he was one hundred per cent sure that the first
man was naturalised. The second man, according to the information that he had,
was naturalised. The third man, according to the indirect information he had,
was naturalised and he further stated that he would not dare to have returned
without a British passport. He did not speak with the second or the third man
because he did not know them personally. They had all safely returned to
Eritrea
for visits and had been able return to the UK. They are all very active members of the
government of
Eritrea
. He is not aware whether they had the permission of the
government to travel here, but the government sends people here amongst the
diaspora. The government plants supporters everywhere. He had met one through
a cousin, the other two he had not met directly. All three went just for
visits. He had never heard of Eritreans going back for good.
84.
People
could be granted refugee status without ever having a dispute with the
government and indeed could be pro-government. He said that the Tribunal in MA
noted evidence that the Eritrean state had an incentive to send supporters to
the diaspora; the government has an incentive to plant its supporters
everywhere. But he had many friends who fled national service in Eritrea
and applied for asylum.
85.
PK
was asked by the bench what percentage of the diaspora in the UK was pro and what percentage was anti the government. He said that this was mysterious and he
could not give an answer. When those who oppose the government have parties
the attendance is small but there is greater attendance at pro government
parties. However, from this one cannot conclude that they are all supporters of
the government. Even though you could not always tell by who attended
different events, only a few supported the government. PK was asked by the bench
whether those going back would necessarily be naturalised; could they return if
for example they had residence in the UK short of naturalisation. He said he
had no evidence about this although he had heard of one man related to a nephew
of his who had died in the UK and his wife wanted to go to Eritrea
where there
was to be his funeral, but she was advised not to go. People would need a
secure status in the country where they live before considering return.
86. PK said he did not know what attitude the Eritrean authorities took to those who returned on a foreign passport; the Eritrean authorities considered them as Eritrean regardless. Asked if such people were not taking a chance returning, he said the overwhelming majority would not take such a risk.
87.
PK
was asked by the bench why people returning to Eritrea
for holidays, even if
they travelled on foreign passports following naturalisation, would risk going
back to a country which he, PK, had described as arbitrary and vindictive. He
said that opinions are personal. Eritrean society is not homogenous and that
there are different political, ethnic and religious groups. Some people may say
his views are unreasonable and others may say the human rights record is
appalling. They think that everything that is going on there is bad, but they
want to take their children back. They are not politically active. As long as
they are not politically active he did not think that the government would go
out of its way because it targets people who undermine them, some would be
pro-government; there may be others who think that as long as they are not
politically active the government would not target them. Mr Knafler asked PK
if he stood by his statement that returning members of the diaspora were only
at risk of being required to do national service if they stayed for a year or
more. He said he did. He knew people who had travelled out from
Eritrea
to Kenya or Sudan to make sure they had not stayed beyond a year.
88.
PK
was asked if the children of those who had fled Eritrea
during the war of
independence would not be considered eligible for national service (on the
basis that they would not be exempt) and his evidence was that the government
targets evaders. He said he did not think the government would target people in
this category. He said that to be fair to the government if you grew up here
and have a passport, in practice national service does not apply to you. He
said that they would be one hundred per cent safe. The government would not be
arbitrary in relation to them.
89.
It
was suggested that in his evidence not many successful asylum seekers would be
returning because they would be deserters/evaders and illegal exiters. His
evidence was that they would be taking a very high risk if they returned.
Successful asylum seekers are not likely to want to approach the Eritrean
embassy, but PK stated that “we can’t talk in absolute terms” there are those
who claimed asylum but have lost parents and now the question of inheritance
arises and they would need to produce documents. It was put to him that
successful asylum seekers would be considered anti–regime, but he said that he
would not go that far and on the contrary and having fled Eritrea
may not be a political.
90. He was asked whether in his view was it not the case that the most typical claim was based on rejection of the regime. He said that once a person gets status here he becomes preoccupied with working and keeping his family. Very few engage in politics. He was asked why, if he accepts that a typical asylum claimant is someone who rejects the regime, why they would not be in the large majority of Eritreans in the UK. PK stated that whether they act to bring about political change is another thing.
91. He did not accept what was said by the Minister of Foreign Affairs who was interviewed in the DFFM who had not indicated that naturalisation is a prerequisite for safe return and who asserted that those who illegally exited were not punished. In respect of evaders/deserters PK found it difficult to accept what was said by the Minister of Foreign Affairs. He stated that this Ministry is not responsible for policy relating to national service in any event. He stated that there may be a “policy” but he had never seen it. PK referred to the immigration officials having said that leaving the country (whilst subject to national service) is illegal. It follows that illegal exit would be subject to punishment which contradicts what was said by the Minister.
92. PK said that he strongly disputed the evidence given by diplomatic representatives, (meeting with diplomatic sources B, C and D) who stated that anecdotal evidence from the US Embassy is that 85 per cent of Americans of Eritrean descent travel on an Eritrean passport. Returnees have to show an Eritrean ID card and in practice returnees show both passports, their Eritrean one and the one they have obtained through naturalisation. The effect of a person using a non-Eritrean passport was that the authorities do not check whether they left illegally.
93.
PK
was asked if he agreed the distinction the respondent seeks to draw between
Eritreans returning having been outside Eritrea
for more or less than three
years. PK said he was aware that persons who go back to
Eritrea
who stayed more than one year were required to do national service, but he knew of no three
year rule of the kind referred to in the UKFFM. Most members of the diaspora
going back for holidays were going for vacations of two-three up to a maximum
of five weeks during the summer. There was another group of retired people who
tend to go back for longer periods; his own brother went back for two months.
He thought a large number of holidaymakers were those who had left
Eritrea
during the war and their children. PK was asked about a paper he had given to EASO
in 2014 where he estimated the proportion of the population who had gone into national
service over a period of 20 years as 9.2 per cent of the population. He was not
the only one who gave high figures, for example a German source had estimated
600,000 people.
94.
PK
was asked to what extent his “dense network” of contacts inside Eritrea
comprised family and friends. He indicated that friends and family comprised a
“small part” and the network spreads across diverse groups. He replied that he
had sources outside the family and inside his family some were pro-government.
95.
PK
was asked questions by Ms Dubinsky relating to his understanding of the present
legal provisions in force in Eritrea
regarding punishments of desertion and
draft evasion with reference to the penal code of the state of
Eritrea
2015 Articles
118 and 119 (inciting to mutiny which carries a sentence of not less than 13
years and not more than 16 and interference with military service which carries
a sentence of not less than one year and not more than three). PK confirmed
that it was not implemented, but said, it did not matter which provision of
national law they applied because victims would never be produced before a
court; they were just put in detention with no recourse. The laws were not
implemented at all. Arbitrariness prevailed and it was impossible to predict
what the State would do, except that punishment would be ”rigorous” and that
could mean detention in containers, in high temperatures or underground cells
and could mean death, torture, indefinite incarceration.
96.
PK
was asked about forcible returns and said that despite inquires he had made, he
only knew of only one forced return to Eritrea
since 2011 and referred to the
case of (Berhane Embaye cited in his Report of 4 April 2016). He was asked why
it was he had little information about forced returns generally and he
suggested two reasons; first that the government is secretive and secondly
because of
Eritrea
’s poor human rights record and people were likely to be detained
if they returned and so most governments were reluctant to deport Eritreans
back to
Eritrea
. It was almost the norm not to. In the last decade it has
generally been accepted that returns would result in ill-treatment. He was of
the view that the second reason was more likely to reflect the position. He had
tried to find out about other deportations but had been successful. He stated
that there had been to deportations from Sudan
97. PK said that should a person be forcibly returned whether they do military or civilian service depended on their skills and what suited the government. In reality even the person sent to civilian national service belonged to a platoon or battalion unit. He could only guess what assignments might be made.
98. He had said the September CIG was not substantially different from the March CIG.
99.
Asked
about the level of surveillance in Eritrea
, PK said it was heavy. Asked if he
thought his relatives there might be kept under surveillance or treated
differently by the regime because of him, he said he did not think the regime
saw him as important enough; they ignored his books. It was suggested that by
any account he must be embarrassing to the regime. He said that the government
ignored these things and that one of his books was in their library. In any
event they did not detain family members except in exceptional cases. His
relatives were safe. He was not politically active. He was not concerned that
their welfare was affected by their relationship with him. Eritrean families
are often divided in their opinion, and you can find within the one family some
pro-government, some anti, some apolitical. He had many friends who were
members of the government, but he did not talk with them about each other’s
political views. He had a brother who was pro-government but they were very
close nonetheless. PK said that he was not important or sufficiently high
profile to be a threat to the government and to cause a problem for his
family.
100. PK
was asked whether he considered the system of national service in Eritrea
amounted to forced labour. He said that when it was introduced national service was
innovative and the national service system was at the heart of the ability of
the independence movement to become a separate state. It was a way of getting
people from different ethnic groups and occupations to come together and put
aside differences and had strong core values that were the building blocks for
the new country. But it all went wrong and once the regime began extending the national
service beyond 18 months problems developed and it became forced labour.
National service for the first 18 months is very much supported by the
populations, but beyond that it is seen as forcing people to do something
against their will. It has degenerated and the fact that military commanders
use national service labour to build their own homes and to recruit labour to their
private projects, illustrates the abuse. Asked to what extent people who did national
service did so voluntarily, PK said that he could not say. People were not
against national service but they were against punishment and against the fact
that it was open-ended. There was a lot of abuse. He accepted that there were
examples of women choosing to stay in the people’s militia to ensure they had a
wage, but that was not a general rule.
101. PK
was asked by Mr Knafler if he had any reservations about fact-finding missions
in Eritrea
. He had said yes there was a risk of round-tripping and because of
the general lack of freedom of expression. But if such missions followed
correct methodology their findings could have value although there were
constraints. In
Eritrea there are so many concerns about getting reliable
information, leaving aside diplomats because it is a closed society.
Schedule of Background Evidence
*Hyperlinks have been included where furnished by the parties