Professor Steve Peers, Royal Holloway University of London
Photo credit: viewsridge,
via Wikimedia
commons
Introduction
Along with judgments on Afghan
women asylum seekers (discussed here)
and the status of Turkey as a ‘safe third country’ (discussed here),
the Court of Justice of the European Union recently gave its first
judgment interpreting the substance of the concept of ‘safe countries of
origin’ in EU law. The judgment addressed a number of important points, interpreting
the law in ways that may have broad relevance to EU law on asylum procedures
going forward.
EU legal framework
Previous rules
The concept of ‘safe countries of
origin’ goes back, at EU-wide level, initially to ‘soft law’ adopted in the
early 1990s (one of the ‘London
Resolutions’ of 1992). Next, the principle first took on binding legal form
at EU level in the first-phase 2005 asylum procedures Directive, which provided
for an option for Member States to
accelerate considering asylum applications (albeit in accordance with the
usual procedural rules), inter alia where the applicant is from a ‘safe
country of origin’. The latter concept was then further
defined: it applied to countries (as described in Annex
II to the Directive) where:
on the basis
of the legal situation, the application of the law within a democratic system
and the general political circumstances, it can be shown that there is
generally and consistently no persecution as defined in Article 9 of Directive
2004/83/EC [the first-phase qualification Directive], no torture or inhuman
or degrading treatment or punishment and no threat by reason of indiscriminate
violence in situations of international or internal armed conflict.
When ‘making this assessment’,
Member States had to take account, ‘inter alia, of the extent to which
protection is provided against persecution or mistreatment by’:
- the relevant laws and regulations of the country and the manner in which they are applied; |
|
|
- observance of the rights and
freedoms laid down in the European Convention for the Protection of Human
Rights and Fundamental Freedoms and/or the International Covenant for Civil
and Political Rights and/or the Convention against Torture, in particular the
rights from which derogation cannot be made under Article 15(2) of the said
European Convention; |
|
- respect of the non-refoulement
principle according to the Geneva Convention; |
|
- provision for a system of
effective remedies against violations of these rights and freedoms. |
When applying these tests, Member
States had to ‘have regard to the legal situation, the application of the law
and the general political circumstances in the third country concerned’, and
the ‘assessment of whether a country is a safe country of origin’ had to ‘be
based on a range of sources of information, including in particular information
from other Member States, the UNHCR, the Council of Europe and other relevant
international organisations’.
Furthermore, the Directive expressly
provided that Member States could designate ‘part of a country as safe
where the conditions in Annex II are fulfilled in relation to that part’. It
was also open to Member States to continue in place pre-existing rules that
fell short of the standards in the Directive to some extent; these rules could
also be applied to part of a country, or to designate a country (or part of it)
‘as safe for a specified group of persons’. There were safeguards
for individuals: for the ‘safe country of origin’ concept to apply, they
would have to be nationals of the country concerned (or stateless persons who
were habitual residents of it), and must not have ‘submitted any serious
grounds for considering the country not to be a safe country of origin in
his/her particular circumstances and in terms of his/her qualification as a
refugee in accordance with’ the qualification Directive. Also, Member States were
obliged to ‘lay down in national legislation further rules and modalities for
the application of the safe country of origin concept’ – although there was no indication
of what those rules and modalities might concern.
Finally, there was a power
for the EU to adopt common lists of ‘safe countries of origin’, by means of
a special procedure. But this power was soon annulled
by the CJEU, when the European Parliament successfully challenged the prospect
of using such a special procedure.
Current rules
The current asylum procedures
Directive, adopted in 2013, retains
the ‘safe countries of origin’ concept as an option for Member States to apply
an accelerated procedure to. There is no longer a reference to potentially treating
only part of a country as safe, and the option for Member States to retain
pre-existing lower standards on this issue (along with pre-existing rules on designating
part of a country as ‘safe’, or as ‘safe’ for groups of people) was dropped. The
notion of a common EU list was not revived in the Directive; although the
Commission subsequently proposed
such a list (consisting of Western Balkans states and Turkey) in 2015, that
proposal did not receive enough support and was withdrawn
in 2019. An express requirement for regular review of the supposedly ‘safe countries
of origin’ by Member States has been added. Otherwise the rules on ‘safe country
of origin’, including the core
definition of the concept, were not amended.
However, the rules on appeals,
elaborated compared to the previous
Directive, now address the issue too. Although the 2013 Directive provides
that in principle asylum-seekers can remain on the territory pending resolution
of their appeal, there is an exception for (inter alia) most accelerated
procedures, including the ‘safe country of origin’ rule. In such cases Member States,
if they opt not to give an appeal suspensive effect, must at least allow a
court to rule on whether or not the asylum-seeker can stay pending the appeal;
and the asylum-seeker can at least stay until a court has ruled on this point.
Before the recent judgment, the
CJEU has only ruled
on these provisions once, and on a purely procedural point: according to
the Court, Sweden could not apply a ‘safe country of origin’ rule until it
legislated for one in national law. Sweden duly did so. In passing, the Court
confirmed the usual understanding that the ‘safe country of origin’ rules set
out a ‘rebuttable presumption’.
Future rules
The 2024 asylum procedures
Regulation has amended
the ‘safe country of origin’ rules again. This entails: explaining the concept of
‘non-refoulement’; confirming that rebuttal of the presumption by an
asylum-seeker must take place ‘in the framework of an individual assessment’; again
expressly providing for exceptions for ‘specific parts’ of the non-EU country’s
territory and (not only in the context of pre-existing law) for ‘clearly
identifiable categories of persons’; and again allowing for the adoption of a common EU
list (this time not via a special procedure, so not repeating the previous
legal problem with the concept of such a list). There is no longer any
reference to national rules providing for further ‘rules and modalities’ of the
concept.
In the broader context, the ‘safe
country of origin’ concept is still on the list of possible accelerated procedures,
but the use of those procedures will now be mandatory for Member States.
In practice the concept might cross over sometimes with the new ground of accelerated
procedures where the country of origin has an international protection recognition
rate below 20%. It will be expressly possible (although arguably it will only
be optional) to apply this ground for accelerated procedures to unaccompanied
minors. The implications of accelerated procedures are more explicitly
harmonised: three months (instead of the usual six months) to decide on the
application, and between five and ten days (instead of the usual two weeks
to a month) to bring an
appeal. The exception to the right to remain pending appeal (subject to a
court ruling otherwise) in such cases becomes
mandatory, with some harmonised rules added: a time limit of at least five
days to request a court to remain on the territory, along with interpretation
and legal aid. As before, accelerated procedures are linked to the rules on a border
procedure, which have also been heavily amended.
National practice
The EU asylum agency has produced
a useful
map of which countries have been designated as ‘safe countries of origin’
(and ‘safe third countries’) by Member States. Furthermore the agency reported
on the national application of these concepts at the end of 2022. As regards ‘safe
country of origin’, the report indicates that a large majority (although not
quite all) Member States apply the concept. The exceptions are Poland, Latvia,
Spain, Bulgaria, Lithuania, Portugal and Romania; Finland applies the concept,
but without a list.
Several Member States apply
geographic exceptions, namely for parts of Armenia, Bosnia and Herzegovina, Georgia,
Moldova, the USA and India. Some Member States also apply exceptions for groups
of people, in particular LGBT applicants, minorities, criminal cases, political
activists, journalists, human rights defenders, victims of discrimination or
forced marriage, and women and girls. There is some variation in which
countries are listed as ‘safe’, with Western Balkans states being listed most
frequently. The report details a number of judgments of national courts concerning
both the listing of particular countries and their application to individual
circumstances, with mixed results – some judgments uphold the listing or its
application to an individual asylum-seeker, but some do not. The recent CJEU judgment is, as we shall see, relevant to the geographic exceptions; and a pending case asks the CJEU about the exceptions for groups of people.
The judgment
The recent judgment concerned the
Czech designation of part of Moldova (except the Russian-occupied Transnistria)
as ‘safe’. Faced with an appeal by a Moldovan asylum seeker who was subjected
to a ‘safe country of origin’ rule, the national court asked three questions:
a) could a country be designated a ‘safe country of origin’ even if it had
triggered the derogation from (some) ECHR rights set out in Article 15 ECHR (which
Moldova had done after the Russian invasion of neighbouring Ukraine)?; b) could
only part of a country be designated as ‘safe’; and c) could a national court
raise the legality of the designation of its own motion?
On the first point, the Court
ruled that invoking Article 15 ECHR did not automatically prevent or cancel the
designation of a country as a ‘safe country of origin’. This was because the
ECHR set out guarantees for use of the derogation in Article 15, limiting its
use to cases where it was necessary, requiring no conflict with other
international law, allowing derogations only from certain rights and remaining
subject to review by the European Court of Human Rights. (On the relevant case
law, see that Court’s guide
to Article 15). Also, it could not necessarily be assumed that invoking Article
15 meant that rights were actually derogated from; nor did use of the derogation
as such determine the nature and extent of the derogation.
However, the CJEU pointed out
that triggering an Article 15 ECHR derogation must mean that Member States had
to review the ‘safe country of origin’ listing, on the basis of the obligation
to review the designations in light of developments set out in the Directive. That
review should be triggered by significant events likely to affect whether a
non-EU country still meets the ‘safe country of origin’ criteria, which include
the invocation of a derogation from the ECHR – even though the derogation would
affect rights other than Article 3 ECHR (the ban on torture or other inhuman or
degrading treatment), because Article 15 ECHR does not allow derogation from
Article 3.
The Court did not address the
arguments in the Advocate-General’s
opinion, concerning the protocol
on asylum for EU citizens by analogy. That protocol provides that EU Member
States are ‘safe countries of origin’ for each other’s citizens, meaning (in
the context of the protocol) that asylum applications from EU citizens cannot be considered at
all – unless a Member State either: invokes a derogation under Article 15 ECHR;
or the Article
7 TEU process of either sanctioning a Member State for a ‘serious and persistent
breach’ of EU values, or warning a Member State that there is a ‘clear risk of
a serious breach’ of such values, has been applied (or triggered, as regards a
warning; on this point, see the Hungary
v European Parliament judgment); or a Member State decides to do so unilaterally
for an individual applicant. In the Advocate-General’s view, the protocol could
not apply by analogy due to its different context: an expression of mutual
trust applicable to EU citizens only.
On the second question – whether
a non-EU country could be designated as a ‘safe country or origin’ only in part
– the Court first looked at the wording of the Directive, which used the word ‘country’
without indicating that it may be referring to only part of the territory.
Secondly, the context of the Directive included the definition of ‘safe country
of origin’ in the Annex; that definition did not suggest either that only part
of a country could be designated, and the words ‘generally and uniformly’
suggested the opposite.
Next, the ‘safe country of origin’
rule was an ‘exception’ to the normal process of considering asylum applications,
and such derogations should have a ‘strict interpretation’ (referring to the
case law on inadmissible applications), notably where an interpretation had no
support from the text. Also, the Court looked at the historical context, noting
that: the 2005 Directive expressly provided for parts of countries to be listed;
the 2013 Directive does not; and the proposal
for what became the 2013 Directive explicitly stated an intention to remove the
possibility. Finally, the Court referred to the objectives of the asylum
procedures Directive, taking the view that the EU legislature was trying to
balance the objectives of a speedy versus a thorough assessment of asylum
applications. This was a discretionary choice that was reversed by the 2024 Regulation;
a change of mind was the prerogative of the EU legislature, provided that it
complied with the Refugee Convention and the Charter. And the wording of the
2024 Regulation reinforced the point that designation of only part of a country
was not provided for in the 2013 Directive.
Again the Court’s judgment does
not discuss a point analysed by the Advocate-General – as to whether the ‘effective
territorial control’ principle, which limits the liability of countries
(including Moldova, in this context) for breaches of the ECHR on their territory
controlled by another State, applies in the context of EU asylum law to allow for
the designation of only part of their territory as ‘safe’. In the Advocate-General’s
view, the issue of state liability for breaches of international law is distinct
from the issue of the listing of ‘safe country of origin’ under EU asylum law;
and the EU legislature decided to provide for listing of whole countries only,
even being aware of the ‘effective territorial control’ issue.
On the third point –
consideration of the designation of ‘safe country of origin’ by a court on
appeal, of its own motion – the CJEU ruled that, taking account of its prior
case law giving a strong role to courts reviewing asylum decisions, including
on procedural issues, such designations have to be reviewed by courts of their
own motion even if the asylum applicant does not raise the point. In
particular, the courts have to review the invocation of derogations from the
ECHR and the territorial scope of the ‘safe country of origin’ designation. Again,
the Court did not address an issue raised by the Advocate-General – the analysis,
in the alternative, of how the EU law principle of equal and effective remedies
would apply in a case like this one.
Comments
The Court’s judgment raises a number of points that could be relevant to EU asylum procedures law more generally. On the first point (the ECHR derogation), the ruling at first sight may only apply to ECHR parties; but logically it applies by analogy also to any derogations from other human rights treaties (including the ICCPR, which is expressly mentioned in the Directive), and/or from national human rights protection – taking into account any relevant differences from the ECHR (the prospects of effective judicial review of the derogation, and the particular rights being derogated from, which were key features of the Court’s judgment). After all, human rights abuses (triggering asylum claims) often take place during supposed states of emergency.
The judgment also makes
clear that reviews of ‘safety’ must be triggered by significant events – not only,
as Member States might prefer, taking place at regular prescheduled intervals. It
follows that a failure to conduct a review in light of a significant event is a
procedural flaw that can be challenged. Moreover, this judgment should logically
apply by analogy to the reviews of the EU common list provided for in the
asylum pact Regulation – as well as to national and EU reviews of ‘safe third
countries’.
As for the dog that didn’t bark –
the comparison with the protocol on asylum for EU citizens – the better
argument than in the Advocate-General’s opinion is surely that in addition to
the obvious difference in wording, the context of the Directive is hugely different too. Although
the protocol uses the phrase ‘safe countries of origin’, unlike the Directive it
does not create a rebuttable presumption, but rather a complete ban on considering
applications (in principle) from EU citizens; so it makes sense that the conditions for applying
the protocol are more stringent, since the key safeguard of the possibility of
the individual right to attempt to rebut the presumption is absent.
On the second point – the
inability to label part of a country as ‘safe’ – as we can see from the practice
above, this would mean changes in a number of Member States. Of course, this is
only short term, because the asylum pact legislation will soon allow such
designations again. But the Court’s express reference to the new Regulation may
give us a broader indication of how it will interpret the pact. It suggests
that while the EU legislature has a discretion to choose between prioritising
speedy or thorough consideration of applications, that choice must comply with the
Charter and the Refugee Convention. Given the reduction in standards in the
Regulation – and the further reductions in standards that the Commission and
many Member States seem to be gagging to make as soon as possible – the Court’s assessment of whether EU asylum law indeed complies
with the Charter and the Refugee Convention is likely to be of increasing
importance in the future.
This part of the judgment also confirms
that the concept of ‘safe country of origin’ is an exception which must be
interpreted strictly, by analogy with the case law on inadmissibility which often
finds that Member States’ interpretations of the grounds for inadmissibility have
been too restrictive. This logically applies by analogy to all grounds for accelerated
procedures, a fortiori after they become mandatory under the 2024 asylum
pact.
Finally, the third point – judicial
review of the ‘safe country of origin’ designation, on its own initiative – logically
applies by analogy too, to other aspects of the principle besides those at
issue in this case, and to listings of ‘safe third countries’ also. Although strengthening
judicial review may have limited effect in such cases, given the absence of suspensive
effect of appeals, confirming the wider power of courts to examine the validity
of designating countries in the list on their own motion may make it easier to
convince them to grant such suspensive effect, given the wider array of remedies which an applicant can therefore seek.