South Sudan’s parliament has now formalized, twice over, a change to how the country will get to its 22 December 2026 election. On 1 July 2026, it ratified amendments to Articles 1.2.5 and 1.2.14 of the Revitalized Agreement on the Resolution of the Conflict in South Sudan (R-ARCSS), removing the national population census and the permanent constitution-making process as requirements that had to be completed before the vote. On 1 September 2026, it went further and wrote the same removal into the Transitional Constitution itself, passing the Transitional Constitution of South Sudan, 2011 (Amendment) No. 13 Bill 2026 in a single day, unanimously, and after ordering journalists and members of the public out of the chamber. [Sources: Radio Tamazuj, “Parliament passes constitutional amendments,” 1 September 2026; “Media barred as parliament debates constitutional changes,” 1 September 2026]
The government’s case is straightforward and deserves to be stated before examining its implications. A census and a full constitution-making process could each take well over a year to complete properly. Removing them as pre-election requirements clears voter registration to proceed within two to three months rather than waiting on processes that risked pushing the vote past its legal deadline again, after elections had already been postponed repeatedly since 2015. Deputy Justice and Constitutional Affairs Minister Joseph Malek defended the July amendments in these terms, arguing that South Sudan has followed the required procedures and has the sovereign right to proceed. Malek was removed from that post at the end of August, in a separate dispute over the boundaries of Abyei unrelated to the elections timetable; the September constitutional bill was carried through parliament by his successor’s ministry, under Justice and Constitutional Affairs Minister Wek Mamer Kuol.
The second amendment matters analytically because it changes what kind of commitment has been broken. A peace-agreement clause can, in principle, be revisited in a future political settlement. A constitutional provision is a different order of legal instrument, and removing the census and constitution-making requirements from the Transitional Constitution itself closes off the argument that the delinking was only ever a temporary feature of the R-ARCSS text. Whatever happens next, NAS and SSOMA can no longer treat the requirement as dormant within the peace agreement; it has been removed from the country’s basic law as well.
Read only as a scheduling decision, the government’s argument has some force. Elections have slipped five times already, and a government that keeps citing incomplete prerequisites as grounds for delay eventually loses the credibility to invoke them again. But the amendments do more than move a deadline. They move the permanent constitution, the legal instrument through which the parties committed to pursue a federal system, out of the pre-election period entirely and leave its completion to a government elected under the existing constitutional framework.
That distinction matters because federalism was never meant to arrive as a political promise. The R-ARCSS links the question of federalism to the permanent constitution-making process. Article 6.2.2 commits the parties to pursue a federal system of governance, while the permanent constitution is the process through which that broader constitutional settlement is to be established. NAS’s own Federal System Model, an 87-page document published in 2020 with a foreword by Chairman Thomas Cirillo Swaka, similarly treats the permanent constitution as the legal framework for establishing decentralized and accountable government.
This is where the incentive problem sits, and it is worth stating plainly that this is a structural point, not a claim about any individual’s motives. A government elected under South Sudan’s current, centralized transitional framework inherits the powers that framework already grants it, now under a fixed five-year presidential term set by the September amendment. Writing a federal constitution afterward would mean voluntarily redistributing authority away from the office that government has spent an election campaign trying to win. Political institutions rarely dismantle arrangements from which those in power derive their authority, not necessarily out of bad faith, but because nothing in the incentive structure pushes them to. The earlier sequence, constitution before election, existed precisely to settle the rules before anyone elected under them had a stake in keeping those rules as they are.
Neither amendment passed without friction, though the friction ran along different lines. The July amendment drew criticism over whether it followed the procedures the R-ARCSS itself prescribes, and the SPLM/A-IO objected that the changes were introduced unilaterally and undermined the agreement’s power-sharing arrangements. A separate proposal, to remove the R-ARCSS’s own legal supremacy over domestic law, was dropped after sustained pressure from international partners, so the agreement’s overall legal framework remains in place even as these provisions have been amended. In September, the friction was internal to parliament rather than between government and opposition: The Legislation and Justice Committee rejected thirteen cabinet proposals that would have cut the size of the national legislature, retaining articles the cabinet had wanted deleted outright, even as the census and constitution provisions passed intact and unopposed. That distinction is worth holding onto: this was not a wholesale abandonment of scrutiny, but a selective one, which makes the outcome harder to dismiss as pure bad faith and easier to describe accurately as a deferral pushed through without an attached safeguard.
That absence of a safeguard is the practical consequence worth naming. Neither amendment appears to attach to the deferred constitution-making process a new, equivalent legal deadline or enforcement mechanism. The election date itself carried a statutory requirement to be announced six months in advance, which the National Elections Commission met in June, and is now fixed at 22 December 2026. The deferred constitution and census carry no equivalent obligation. A commitment with no enforcement date attached is not automatically abandoned, but it is not automatically kept either, and the difference between the two will depend on pressure that, at present, has no legal trigger forcing it.
The September vote adds a procedural footnote to that same absence of safeguard. The bill that wrote the delinking into the constitution was itself passed after the Speaker ordered the media and the public out of the chamber, citing national security and the need to protect sovereignty, public order and social stability. If the government’s case for the amendments rests on due process having been followed, as Malek argued of the July changes, then the process delivering that outcome should reasonably be visible to the public it binds. A closed sitting is a difficult basis on which to ask sceptical observers to trust the process, whatever the substantive merits of removing the census and constitution as pre-election requirements.
NAS’s argument is more specific than a call to wait for conditions to improve. It makes a specific causal claim: that elections held without political consensus and functioning safeguards would not unite the country but would instead accelerate political fragmentation, insecurity and institutional collapse. On this view, a vote conducted under the current constitutional and security conditions does nothing to resolve the conflict driving the crisis. It hands one side within that conflict a formal mandate to govern while leaving the grievances of factions who reject the outcome exactly where they were. That is a different prediction from the government’s, which treats the vote itself as a step out of the transition rather than a trigger for continuing it. The disagreement, therefore, is not simply over whether elections should occur, but over what elections are expected to accomplish.
SSOMA’s position adds an institutional dimension to the same argument. In its February 2026 response to the Tumaini framework, the alliance argued that sustainable peace requires a process addressing governance failure, security sector reform and justice together, rather than a settlement that rearranges the existing agreement or defers those reforms to a later date. Read alongside NAS’s statement, what the two movements are proposing is not an indefinite postponement but a different order of operations: build the institutions capable of managing conflict first, then hold a vote those institutions can absorb, rather than holding the vote first and leaving the institutions to catch up afterward.
The concern is not confined to the opposition. During an African Union field mission in April 2026, the Troika, the European Union, the IGAD Partners Forum, civil society and faith-based organizations independently warned that a unilateral electoral process risked returning the country to full-scale war, a warning echoed in the UN mission’s own reporting on unresolved security arrangements and shrinking civic space ahead of December. That is meaningful corroboration from outside the opposition’s own ranks. At the same time, those same actors have warned against an open-ended transition with no electoral endpoint at all, a position the AU’s own C5 committee has taken as well. The external position is therefore more complicated than either side’s preferred formulation: there is pressure to hold elections, but also recognition that elections held without the necessary political and institutional conditions risk deepening the crisis rather than resolving it.
There is a reasonable counterargument here, and it deserves a direct answer rather than a dismissal. A constitution written by an elected government could claim a democratic mandate that a transitional process cannot, and an elected legislature may have a stronger democratic claim to adopt a permanent constitution than institutions operating under a peace agreement. On that view, sequencing the vote first strengthens the eventual constitution’s legitimacy rather than weakening it. The difficulty is that a mandate to govern under the existing rules is not the same as a mandate to rewrite those rules against the interest of the office that mandate was given for. An election settles who holds power under the current framework. It does not necessarily settle what the permanent constitutional order should be, and it does not, by itself, generate any pressure to change that framework afterward.
One attribution point is worth being precise about, and it now applies twice over. NAS’s 4 July statement lists the absence of a permanent constitution and a credible population census among its stated conditions for credible elections, but it addresses the election process broadly and was not framed as a direct response to the 1 July amendment, which it predates by three days. It obviously cannot be read as a response to the 1 September constitutional amendment either. What both amendments do is give NAS’s existing argument a sharper point of reference: the constitutional process that was expected to precede the election has now been formally moved beyond it, twice, in both the peace agreement and the constitution itself. Any future NAS or SSOMA statement addressing the September amendment specifically should be sourced and dated on its own terms rather than assumed from the July position.
What would make the deferral credible rather than indefinite is not difficult to identify. A binding timeline for the constitution-making process, independent oversight of the body that drafts it, and continued monitoring by the African Union and IGAD would each narrow the gap between a genuine postponement and a quiet abandonment. That gap has, if anything, widened rather than narrowed: the requirement is now absent from both the peace agreement and the constitution, with no equivalent obligation written into either in its place. Absent those safeguards, the question is no longer simply when South Sudan will get a permanent constitution. It is who will write it, and whether the government that wins power under the existing centralized system will have any incentive to give that power away.
The writer is a lawyer, political activist, and a member of the National Salvation Front (NAS). She can be reached via tereza89@yahoo.com.
The views expressed in ‘opinion’ articles published by Radio Tamazuj are solely those of the writer. The veracity of any claims made is the responsibility of the author, not Radio Tamazuj.
and then