Aly Verjee 24/12/17
Later this evening, a cessation of hostilities agreement (CoHA)
between the government of South Sudan and eight South Sudanese armed
groups comes into effect. The agreement, signed in Addis Ababa and
mediated by IGAD, was welcomed by many South Sudanese and the
international community (see statements from the AU, Troika, and EU).
The African Union Chairperson called the CoHA a ‘critical first step
in the efforts to end the senseless conflict and carnage that has been
unfolding in South Sudan since December 2013.’ This CoHA, is,
however, the eighth agreement since January 2014 to speak of ending
While I’d like to be hopeful about this agreement and join in the
optimism of these statements, this CoHA is entirely conventional. It
is thus likely to be dogged by the same problems that have seen past
All cessation of hostilities and ceasefire agreements depend on three
factors to succeed:
the will to implement;
the ability to investigate, verify and deter violations;
and, the means to avoid escalating minor breaches of the ceasefire, so
that, for example, one undisciplined soldier firing a rifle doesn’t
cause a full scale battle.
This CoHA falls short on all three counts.
It doesn’t improve the ceasefire monitoring body’s ability to
investigate violations, or how those investigations are acted on once
complete. The ability to deter violations is therefore unchanged.
There have been at least 50 documented violations of the earlier
‘permanent ceasefire’, with little to no follow-up or meaningful
This CoHA doesn’t specify how de-escalation and conflict mitigation
could occur. Article 18 calls for disputes to be ‘resolved amicably’
by the combatants. If amicable resolution was that straightforward,
South Sudan would not be in a multi-dimensional civil war.
And as far as the will to implement goes, the CoHA’s preamble itself
acknowledges the problem: ‘having in mind the lack of genuine will
and robust mechanisms to ensure enforcement and compliance with those
commitments,’ it says, let’s try again. The test is in the days and
weeks ahead. But, in addition to the seven earlier agreements and
re-dedications, should be added the government’s declaration in May
2017 of a unilateral ceasefire, and renewed in November. Yet
offensive action by government forces continued throughout this
period: those actions loom ominously over any text.
Since the facilitators of the IGAD High Level Revitalization Forum
(HLRF) offered a weak and ambiguous draft text at the outset, it’s no
surprise that these three fundamental factors were insufficiently
addressed, even if the final CoHA text has somewhat improved. The
list of twenty concerns that follows is not aimed at disparaging the
efforts of my former colleagues. It is, instead, intended to identify
the weaknesses of the text and suggest what should be addressed in the
next phase. While the road to peace is long, we make the journey
harder by failing to address ambiguities and avoidable
Weaknesses of the December 2017 CoHA:
1. CoHA timelines are inconsistent. Art. 1(1) states the CoH comes in
to effect from 0.01 hours on the morning of 24 December; Art. 1(2)
speaks of an ‘immediate’ freeze of forces in their locations, although
the implementation matrix (p. 21) defines immediate as from 0.01 hours
on the morning of 24 December, and therefore, by definition, something
that is not immediate. Art. 1(4) requires forces be notified of the
terms of the CoHA ‘within 72 hours of the signing.’ Since the
agreement was signed around 2000 hours on 21 December, notification
could be given up to 2000 on 24 December, after the CoH has already
come into effect, and therefore, its implementation is delayed from
2. There is no prioritization of locations in the CoHA. As
complicated as South Sudan’s conflict has become, violence is not
evenly distributed. Some parts of the country face more intractable,
multi-actor conflicts. The CoHA would contribute practically if it
helped prioritize the most difficult areas for conflict resolution and
areas of active hostilities. Without prioritization over a large and
inaccessible geographic area, the investigation of violations and
monitoring of the agreement becomes impractical.
3. Provisions on the movement of forces are ambiguous. Art. 1(2)
requires movements of forces be authorized, with exceptions for
vaguely defined ‘logistical and administrative purposes,’, an
exception open to abuse. More seriously, Art. 11(6) only requires
‘advance notification’ of the movement of forces, which conflicts with
the earlier article and provides the combatants with a loophole that
arguably allows force movement to be less constrained. There is no
time period specified for ‘advance’ notification, which defeats the
intent of the provision. The implementation matrix only notes
notification should be in writing.
4. Provisions for the disengagement of forces are unspecific. Art.
1(3) requires ‘where opposing forces are in close proximity they shall
immediately disengage.’ Close proximity is left undefined. The
firing range of small arms? The firing range of medium and/or heavy
weapons? Forces that are within 1km of each other? 2km? What
constitutes an adequate distance for the withdrawal of forces is also
5. Accountability measures are unspecified. Art. 3(5) requires the
parties ‘ensure accountability for breaches of the agreement,’ as does
Art. 13. There is no indication as to how accountability should be
pursued, or on what time frame: formal reprimand? The commanding
officer being suspended or relieved of command? Court-martial?
Within 72 hours? 7 days? A timeline determined by the offending
6. Withdrawal from civilian installations is unspecific. Art. 4(2)
does not specify what constitutes withdrawal from public institutions
such as schools and hospitals, nor state what steps are can be taken
in the event withdrawal is not completed by the implementation matrix
deadline of 24 December.
7. Opens the door to offensive action on the grounds of the protection
of civilians. Art. 7(1) requires the parties ‘to protect civilians
from any form of attacks by other armed actors not part of this
agreement.’ How, in such circumstances, one protects civilians from
attacks without resorting to offensive action is unclear. Ostensibly
a provision that provides for the protection of civilians is an
invitation to engage in further violence and commit further
8. Disclosure to CTSAMM is problematic. Arts. 2(2) and 11(8) require
the disclosure of sensitive information to CTSAMM, without specifying
to whom at CTSAMM disclosure should be made. In the absence of
confidence building measures, it is unlikely accurate disclosures will
be made openly to the entire CTSAMM board, which includes other
9. Self-defence justification is unproscribed. Self-defence has been
frequently invoked as justification for past ceasefire violations.
This CoHA does not limit the right of self-defence in Art. 3(2), and
leaves it to retrospective determination by CTSAMM as whether
self-defence was validly invoked, in Art. 3(4).
10. Reporting requirements are unchanged. Art. 14 maintains the
current reporting chain in the event of violations: from CTSAMM to
JMEC, and from JMEC to the TGoNU, IGAD, AU Peace and Security Council
and UN Security Council. It is unclear how an identical reporting
protocol will galvanize any new action to be taken in the event of a
11. Media restrictions remain. Art. 5(2)(d) requires the parties to
‘ensure protection of the media,’ and Art. 10(g) prohibits ‘all forms
of harassment of the media’. Both are sound principles, but do not
specifically remove restrictions on media in South Sudan, such as
pre-publication censorship, onerous registration requirements, or
freedom of movement, including the issuance of visas, to international
12. Political prisoners and prisoners of war (POWs) provisions are
inadequate. Art. 9(3) obliges the parties to declare the political
prisoners and POWs in their custody. This requirement is easily met by
a declaration that the party holds no political prisoners or POWs and
that any detainees held face criminal charges. Yet there is evidence
that specific individuals are being held because of their political
beliefs. The agreement could have specifically provided for these
named individuals to be freed. As it stands, this provision will not
cause the parties to change their positions.
13. ‘International best practice’ on the cessation of hostilities and
provision of humanitarian access is unspecified. Art. 2(4) requires
parties to abide by international best practices, without specifying
what such practices are or which are most critical for adherence.
14. ‘Specific’ measures to protect civilians are unspecified. Art.
5(2) requires the parties to take ‘specific measures’ to protect
vulnerable persons, support the reunion of families, facilitate the
free movement of persons, etc., without any indication of what the
appropriate and specific measures to achieve these goals should be.
15. Proximity to protection of civilian (POC) sites, IDP and refugee
settlements unspecified. Art. 6(h) requires the parties not to
‘incit[e] violence in or around’ POC, IDP and refugee sites, without
specifying what constitutes such proximity.
16. Hostile verbal attacks are not proscribed. ‘Unwarranted’ verbal
attacks on the IGAD mediation, ceasefire monitors, JMEC, etc., are
prohibited by Art. 4(1). So, are there such things as warranted
verbal attacks? Minister of Information Michael Makuei will be
17. Procedure for joining the CoHA is unclear. Art. 2(3) calls for
other combatant groups to join this Agreement, without providing the
modalities (timeframe, mechanics) for that to occur.
18. Dispute settlement provisions are inadequate. As mentioned
earlier, Art. 18(1) unrealistically calls for ‘amicable resolution’,
of disputes in the implementation and interpretation of this
Agreement. Art. 18(2) refers unresolved matters to the IGAD Council
of Ministers, a problematic venue as the government of South Sudan, a
combatant group which may be party to dispute about the CoHA, sits at
the Council of Ministers’ table.
19. Amending the CoH is possible without consensus. Although this
agreement was reached by consensus, amendment of the agreement is
possible by a majoritarian decision, as provided by Art. 17. This
leaves the door open to unhelpful actions by a coalition of some
combatant parties against others. Since non-combatant parties are
also included as signatories to this agreement, it is also unclear
which parties count for the purposes of amending the CoHA.
20. Raises modification of JMEC’s composition prematurely. Art. 14(5)
calls for the restructuring of JMEC. This is premature to address in
a CoH agreement and gives parties an opening to argue that JMEC should
not work pending its reconstitution, although the implementation
matrix caveats this provision by stating that this will occur ‘as
decided by the HLRF process’. It would have been better to omit this