Yemen's peace process: the need to change the international vision and framework

Abstract

• Since the start of the Saudi-led military intervention in Yemen in March 2015, the basis for the dominant international understanding of conflict in Yemen, and for the dominant vision for a peace process, has been that Yemen is in a civil war, and that peace requires a return to a sovereign national government in Sana'a and to the political transition arrangements which were in place during 2011–14.

• This vision has tied mediation efforts to an increasingly inappropriate framework and ignored realities about governance in Sana'a, the internationally recognized government and the issues of contention, as well as the need for Riyadh–Sana'a talks.

• As a consequence, opportunities for making greater progress on peace have been missed. The ineffectiveness of the peace process has also contributed to the conditions in which hard-liners in Ansar Allah—the Houthi-led political movement in power in Sana'a and northern Yemen—have been able to strengthen their control, reducing the possibility for peaceful political change.

• International policy-makers and practitioners need to change the international vision and framework for Yemen's peace process. They should:

  • Acknowledge realities about the rival governments, recognizing that hybrid and multiple arrangements in governance can be compatible with peace;

  • End inappropriate insistence on UN Security Council Resolution 2216 and allow it to be formally superseded; and

  • Facilitate inter-Yemeni dialogue, to the extent that they can, without trying to direct a process and formulate agreements.

• These and other recommended shifts in approach will ease constraints on progress and widen the space for a Yemeni vision and Yemeni-led peace process to develop.

In September 2014 Houthi armed forces allied with the former president of Yemen—Ali Abdullah Saleh—and with forces loyal to him captured the capital, Sana'a. March 2025 marked the tenth anniversary of the start of the military intervention by the Saudi-led Coalition to Restore Legitimacy in Yemen, which aimed to restore the government ousted by the Houthis, led since 2012 by Abd Rabbu Mansour Hadi. During the past decade the basis for the dominant international understanding of conflict in Yemen, and for the dominant vision for a peace process, has been that Yemen is in a civil war, and that peace requires a return to a sovereign national government in Sana'a and to the political transition arrangements which were in place during 2011–14.

Drawing on open-source material and the author's observations from working in Yemen in an international mediation capacity, this policy paper explores how that vision for a peace process has been unnecessarily and unhelpfully constraining. The vision has meant that mediation efforts led by the United Nations have generally been tied to an increasingly inappropriate framework of references to international resolutions and documents, ignoring realities about governance in Sana'a, the internationally recognized government and the issues of contention, as well as the need for talks between Riyadh and Sana'a. The consequences have been significant. Opportunities for making greater progress on peace have been missed. And the unrealistic vision and inappropriate framework for peacemaking have contributed to the conditions in which hard-liners in Ansar Allah (AA)—the Houthi-led political movement in power in Sana'a and northern Yemen—have been able to strengthen their control, reducing the possibility for peaceful political change.

The paper concludes with recommendations for international policy-makers and practitioners about the changes that are needed in the international vision and framework for Yemen's peace process. Central to these are the need to acknowledge realities about the rival governments, and about current and future governance and peace in Yemen, and to move more towards facilitating—rather than directing—Yemen's path to greater peace.

The vision for a Yemeni peace process and mediation

The international vision and framework for a Yemeni peace process have remained largely static during the decade since the Houthi take-over of Sana'a. This has been despite meagre results and poor alignment with the situation in the country.

The dominant vision for a peace process

Since 2015 the primary international reference point for a peace process in Yemen has been UN Security Council Resolution (UNSCR) 2216, adopted in mid-April 2015, a little under three weeks after the start of the Saudi-led intervention. The associated international mediation has primarily been conducted by special envoys of the UN secretary-general, carrying out the latter's ‘good offices’ role under a mandate which began in 2012 and developed by way of resolutions over subsequent years.¹

UNSCR 2216 called upon ‘all Yemeni parties, in particular the Houthis’ to abide by the 2011 Gulf Cooperation Council (GCC) transition agreement for Yemen, the outcomes of the 2013–14 National Dialogue conference, and ‘the relevant Security Council resolutions’.² Over subsequent years, these three elements (chief among them UNSCR 2216 itself) came to be habitually termed the ‘three references’ by the internationally recognized government of Yemen, as well as some diplomats and allies.³ This insistence, in effect a policy, has continued through to the present time, and has been communicated repeatedly by government officials, supporters and the GCC.⁴ It has thus become a prevailing framework for a peace process for Yemen. Insistence on the ‘three references’ has not prevented critiques of UNSCR 2216: for example, that it imposed constraints which made successful negotiations impossible, that it ignored problems with the meaningfulness of Hadi's claim to legitimacy, and that it effectively sought surrender and disarmament by the Houthis.⁵ The most succinct critique of the resolution is that it had a ‘negative impact on attempts to end the conflict’.⁶

In the early years of the Saudi-led intervention, diplomatic efforts to replace UNSCR 2216 were made but were resisted by some Security Council members, among them the United States and the United Kingdom; the latter was the ‘penholder’ in this period for resolutions on Yemen.⁷ Since then, the growing obsolescence of UNSCR 2216 has sometimes been recognized internationally, but not widely enough to have prompted open acceptance of its obsolescence, or decisive action to replace it.⁸

Meagre results

High-level and top-down peace processes are by their nature difficult and will often yield meagre results. But this is not a good reason to avoid examining results and asking questions about them. For Yemen, in the ten years from 2014 to 2024, high-level peace efforts (the peace process, as it was usually called) yielded only two agreements that were central to the conflict: these were the 2018 Stockholm agreement and the renewable two-month truce agreement reached in April 2022.⁹ The former was a short agreement that essentially provided a ceasefire in the city of Hodeidah, de-escalating and freezing a military front line. The latter was a verbal agreement to a UN truce initiative providing for a temporary halt to major hostilities and for some easing of the sea, air and road obstructions in northern Yemen. Formally, the truce was between Sana'a and the official government, but unofficially it was based on an agreement between Saudi Arabia and Sana'a, with Saudi Arabia pressing the official government to accept it and unveiling, in Riyadh, a new official Yemeni government executive (the Presidential Leadership Council) and president (Rashad al-Alimi), five days after the UN announced the truce.¹⁰

Other proposals and talks either were not carried forward or were inconclusive. Between April and August 2016, for example, UN-mediated talks were held in Kuwait, where a delegation representing AA and the Sana'a section of the General People's Congress (GPC)—at that time led by Saleh, the former president—met with a delegation from the official government led by Hadi, but without reaching an agreement.¹¹ In April 2020 Sana'a issued an eight-page peace agreement proposal, and Riyadh issued a one-page proposal for a temporary ceasefire, neither of which were taken up by the opposing party or carried forward by the UN mediation.¹² Furthermore, in 2020–21 the UN mediation team unsuccessfully pursued a ‘joint declaration’ initiative which was intended to be between the official government and Sana'a.¹³

What the vision got wrong

That the peace process produced only meagre results was largely because of four important factors that the international vision for a Yemeni peace process did not recognize. These were: the realities of the government in Sana'a and its governance; realities about the official, internationally recognized government; realities about issues of contention for negotiations; and the need for talks between Riyadh and Sana'a.

The realities of governance in Sana'a

International narratives about the Houthis, AA and the government they lead in Sana'a have tended to under-represent the degree of power and governance that they exercise. Security Council resolutions have persistently referred to the government in Sana'a as ‘the Houthis’ and have avoided the words ‘Ansar Allah’ and ‘government’ (or the government's name).¹⁴ The public language of western and Arab states has generally replicated this, in some cases with more liberal labelling of the Houthis as a militia.¹⁵ More widely, international media reporting about Yemen has labelled the Houthis variously as militias, rebels, Iran-backed, and terrorists. The official government has assiduously stuck to calling the Houthis ‘militias’ and ‘terrorists’.¹⁶ Meanwhile, UN and international non-governmental aid organizations operating in Sana'a-controlled areas have long used language that acknowledges the reality of the Houthi-led government, talking of the de facto authorities (referred to by the acronym DFA) and of AA, addressing Sana'a government ministries and officials by their titles and, in parallel, referring to the official government as the internationally recognized government (IRG) or the Government of Yemen.¹⁷

Collectively, these terms and circumlocutions have obscured the extent and capabilities of Houthi and AA rule, which have comprised functioning and relatively stable central, provincial and local government and administrative structures, governing territory which is home to around 70 per cent of Yemen's population. Those structures include, at the formal apex, a ten-member Supreme Political Council (SPC), of which five members are from AA and five from the Sana'a-based section of the GPC; a cabinet government, called the Government of Change and Construction (until 2023 the National Salvation Government) with ministers from AA, the GPC and other minor parties; and a parliament that meets and carries out conventional parliamentary activity.¹⁸ SPC-appointed governors preside over administrations in governorates controlled by Sana'a, and nominal SPC-appointed governors also exist for the minority of governorates that Sana'a does not control. Since 2019 the Sana'a government has had a written national vision and strategy extending to 2030, which it publicly pursues, even if the results are scarce.¹⁹ Less explicitly—but more consequentially—AA has an ideological programme which is articulated in the speeches and sermons of the movement's leader, as well as the organization's copious media output, and is applied in policies and actions in many areas of the economy, politics and society.²⁰

Acknowledging that the Sana'a government exists is not to approve or disapprove of its politics or its performance. Indeed, AA and the Sana'a government have undoubtedly acted and governed in repressive, authoritarian and belligerent ways, and continue to do so. But a peace process will not be effective if it is not congruent with the situation that exists in a country. International reporting and academic research on Yemen during the past decade have been hindered by the war and political obstacles. However, there are examples of both types of work that communicate some of the realities of the political and social situation in northern Yemen, as well as the history of Houthi relations with the state and of the series of military conflicts between 2004 and 2010 that became known as the Sa‘dah Wars.²¹ This background of conflict and what Houthi leaders saw as political and economic marginalization created the grievances that, since 2014, have motivated their and AA's political drive to build their power and rule.²² These aspects should be noted, too, when claims are made about Iranian influence over AA. Iranian backing for AA has been real, but its extent and influence over AA's decision-making have tended to be overstated by opponents and outsiders underestimating Houthi and AA motivations.²³

The realities of the official government

Since 2014, the dominant international narrative about Yemen has included an insistence on the legitimacy of the official government, the IRG. This has a basis in the history of central government in Yemen during 2011–14, when a transitional government of national unity was formed under the 2011 GCC transition agreement for Yemen. Since 2015, however, the meaningfulness of the government's status as legitimate and internationally recognized has been severely eroded by time, absence (including some dependence on Saudi Arabia and the United Arab Emirates—UAE), and the limited extent of the governance for which it has actually been responsible within Yemen (where at most one-third of the population lives in territory firmly or nominally under IRG control).

The 2011 GCC transition agreement provided for a transitional government headed by a president who would serve a two-year term until elections were held. Thus, when Hadi became president in an uncontested election in February 2012, he was due to serve until February 2014, though in January 2014 (after delays in the National Dialogue) parliament extended his term by one year, to February 2015.²⁴ In January 2015, when under house arrest in Sana'a following the Houthis’ capture of the capital, Hadi announced his resignation in protest at Houthi control of state institutions. Parliament did not accept his resignation, and in February, after fleeing to Aden, Hadi announced he was revoking his resignation. In March Hadi called on Saudi Arabia and the GCC to intervene against the Houthis and to protect Yemen, invoking the principle of self-defence.²⁵ In April UNSCR 2216 reaffirmed Security Council support for the ‘legitimacy’ of Hadi as president, and without further international measures endorsing his presidency, Hadi retained the presidency for seven more years.²⁶ In April 2022 his tenure abruptly ended when Saudi Arabia and the UAE oversaw the formation in Riyadh of the eight-member Yemeni Presidential Leadership Council, to which Hadi transferred his powers under Saudi and Emirati encouragement—or pressure. The decree which established the council did not set a term for its duration and since its creation the council has struggled to overcome internal divisions and present a united front.²⁷

During this time, the IRG, its ministries and national bodies such as the central bank have formally been headquartered in Aden. In practice, the president and ministers have spent much time outside the country, in particular in Riyadh, where they have had offices and have often received foreign officials. Through diplomatic activity and Yemeni embassies abroad (which remained overwhelmingly under its control), the IRG has encouraged international actors to maintain exclusive recognition of it as the government of Yemen. Similarly, media under the government's control, or supportive of it, have consistently referred to the government in terms that maintain the narrative of legitimacy.

Nonetheless, international legitimacy and insistence upon it are not the same as domestic legitimacy or tangible authority and governance. Yemenis, and international commentary, have rightly questioned the labels of legitimacy and international recognition, as well as the capabilities and conduct of the IRG.²⁸ The government's authority has also been questioned in UN reporting. In 2018 the UN panel tasked with monitoring UN sanctions on Yemen judged that the authority of the IRG had ‘eroded to the point that it is doubtful whether it will ever be able to reunite Yemen as a single country’.²⁹ The IRG's limited authority has also been evident in its relationship with the Southern Transitional Council, which contests the IRG's authority in the south and has governed Aden since 2018, notwithstanding a Saudi and UAE attempt in 2019 to broker an agreement between it and the IRG.³⁰

Realities of issues of contention

The third factor that the dominant international vision for a Yemeni peace process did not adequately recognize related to the issues of contention that needed to be addressed. Instead, narrow adherence to the framework of the ‘three references’ tended to impede the scope, and efforts, to address issues that were either central to the situation of war or were otherwise important to civilian well-being. The former category comprised major military activity (affecting all parties) and aerial and maritime blockades (affecting the Sana'a government and the population under its control). The latter category comprised a range of economic and infrastructural issues affecting all parties and the population in general.

Neither UNSCR 2216 nor any other Security Council resolutions adopted between 2015 and 2022 spoke of the Saudi-led coalition's military campaign, or called for a nationwide ceasefire. Likewise, apart from an oblique reference to Sana'a airport in Resolution 2451, Security Council resolutions did not speak of the closure of Sana'a airport to commercial air traffic after 2016, nor did they deal with the restrictions on marine traffic to and from the ports of Hodeidah, Salif and Ras Issa which were instituted by the IRG and enforced by the Saudi-led coalition (with a donor-funded project, the UN Verification and Inspection Mechanism, set up in 2016 to inspect and issue clearances for ships seeking to enter the ports, and an offshore ‘Coalition Holding Area’ to enforce arrangements).³¹ In the opinion of some observers and human rights rapporteurs, the restrictions on Sana'a airport and the seaports were, or were tantamount to, air and sea blockades.³² Consistently through the war, and since the 2022 truce, AA and the Sana'a government have called the coalition's military campaign the ‘aggression’, and the restrictions on Sana'a airport and the ports of Hodeidah, Salif and Ras Issa a ‘siege’ or ‘blockade’.³³ AA's position—as expressed, for example, in speeches by its leader Abdel-Malik al-Houthi—has been that ‘ending the aggression and the siege’ are humanitarian and economic issues that must be addressed before or with other issues.³⁴

Other issues of contention that a peace process could reasonably seek to address have included the existence of two Yemeni currencies with different values, public sector salary arrears, oil revenue use, cross-lines roads, electricity generation and distribution, and the presence of foreign military forces in some locations. These issues have been among the focuses of international efforts to generate progress in talks, for example during the 2022 truce and, since then, in UN proposals to extend it.³⁵

The scope for international mediation to give due priority to these various issues of contention has, however, been narrowed by the tendency to treat UNSCR 2216 as though it specified the terms for a peace process. It has also been reduced by the insistence of the IRG and its supporters that the ‘three references’ should guide the peace process.

The need for talks between Riyadh and Sana'a

The fourth important factor that the prevailing international vision and framework for a peace process did not recognize was that talks between Riyadh and Sana'a were needed. The fact of the Saudi-led military intervention, the history of Saudi security interests in northern Yemen (particularly Sa‘dah), and AA's clearly held perception that Saudi Arabia was a pivotal actor, always pointed to the possibility of and the need for Riyadh–Sana'a talks. AA were not novices to negotiation, having experienced years of failed mediation and negotiations from a position of inferiority during the Sa‘dah Wars.³⁶ However, UN mediation efforts and Security Council discussions about Yemen spoke only of mediation and talks between the Houthis and the official government.³⁷

AA's perception that Saudi Arabia was a party to the war and an essential party to eventual talks was frequently expressed in its discourse about Saudi ‘aggression’ and was reflected on occasions when sporadic opportunities arose in the early years of the war for contact with Riyadh—for example, in-person talks in 2016 about a ceasefire, at a Saudi military base, and indirect talks over video in late 2019.³⁸ The same perception was signalled in specific proposals, such as the comprehensive solution document proposal of April 2020. This proposal was largely ignored internationally—including by UN mediation—firstly because of a tendency to disregard statements by Sana'a, and secondly because the proposal was presented as an agreement to be signed by two parties, namely the ‘leadership of the Republic of Yemen in Sana'a’, and ‘the leadership of the Coalition’, meaning respectively Sana'a and Riyadh.³⁹ In the subsequent months, AA reiterated its support for this proposal, contrasting it with what a senior AA figure described as the ‘lack of any vision’ proposed from ‘any party from among the aggression countries’ for a solution to the war.⁴⁰

The 2022 truce and the subsequent talks that Oman facilitated between Riyadh and Sana'a were further evidence of the need for such talks. The role of Riyadh in producing the official government's agreement to the truce in April 2022 has not been made public. However, after the truce lapsed in October 2022 Saudi Arabia gradually accepted that it would enter direct talks with Sana'a and would be seen publicly as doing so, and the necessity of Riyadh–Sana'a talks to end the war was acknowledged in some international reporting.⁴¹

Separately from meetings about prisoner exchanges, a Saudi delegation visited Sana'a in December 2022 before making a more public visit in April 2023.⁴² The following September a delegation visited Riyadh from Sana'a to meet the Saudi defence minister Prince Khalid bin Salman, who afterwards publicly referred to the visitors as the ‘Sana'a delegation’—rather than the Houthis or Ansar Allah—but still framed Saudi Arabia's role as being to support the peace process and inter-Yemeni dialogue to reach a ‘comprehensive political solution under UN supervision’.⁴³ Saudi Arabia's hesitancy to be seen as a party to negotiations and the war itself was repeatedly noted by Sana'a in its discourse.⁴⁴

Talks between Riyadh and Sana'a also made sense because of shared security interests and history. That history includes previous significant Saudi-Yemeni agreements (e.g. the 2000 Treaty of Jeddah, and, before that, the 1934 Treaty of Taif), contested border demarcation and fortification, Saudi financial patronage in Yemeni borderland governorates, interaction between Saudi border policy and the emergence of the 2004–2010 Houthi conflict (the Sa‘dah Wars), and tribal displacements and border clashes after 2011.⁴⁵ The mutual border has long been a focus of Saudi-Yemeni discussions, and is naturally a significant factor in maintaining stability between the two countries.⁴⁶

Consequences and recommendations

The shortcomings of the dominant international vision for a peace process for Yemen have had two major consequences for the country. Firstly, they have saddled Yemen with an ineffective peace process that has yielded meagre results because of what the vision ignores, excludes or gets wrong. UNSCR 2216 and the ‘references’ for peace were linked to the Saudi-led military intervention and a goal of restoring the IRG to power. The credibility of that goal was rapidly eroded by the intervention's lack of success. But with foreign diplomats and donor country officials working on Yemen being based almost without exception outside Yemen (either in their home capitals or elsewhere in the Middle East) and rarely visiting Sana'a,⁴⁷ the prevailing understandings of conflict and peace in Yemen have been shaped primarily by regional and international political preferences and interests and not by the realities of the situation. UN offices within Yemen have been constrained in their ability to influence international policy, and the credibility of the UN as a neutral mediator has been eroded by the degree to which its mediation efforts have been constrained by a vision based on UNSCR 2216.

The second major consequence for Yemen has been that, by not addressing what was needed and by producing few results (until the truce in 2022), the nominal peace process has contributed to the time and opportunity that AA and the Houthis have had to consolidate power and become entrenched in uncompromising foreign and domestic policies. The combination of the Saudi-led intervention and an ineffective peace process have unwittingly tended to reinforce the Houthis' development as state actors, contrary to the initial ambitions of the intervention.⁴⁸ AA hard-liners with pessimistic and Manichean beliefs about Israel, Arab and western countries, and the UN, have felt those beliefs to have been proved correct. Moderates inside and outside AA have struggled to counter the hard-liners' beliefs and narratives about regional and international politics and, consequently, have seen what little influence they had on the Sana'a government dwindle.

The signs of this entrenchment have been abundant, from Sana'a government statements and the persistent promotion of the sarkha slogan,⁴⁹ to the silencing of potentially moderating figures in politics and society in Sana'a, and the media output and murals in the streets of Sana'a decrying the Security Council's perceived acquiescence or complicity in the ‘aggression’.⁵⁰ The Israel–Gaza war which began in October 2023, and the lack of pressure from Arab and western countries for a rapid ceasefire, have added to the conviction of AA hard-liners that they are right and they should double down on combative policies and actions, such as drone and missile attacks on Israel and on ships in the Red Sea, and arrests of Yemenis working for the UN and for international and Yemeni NGOs.⁵¹ AA's leaders were also not persuaded by the prospect, at the end of 2023, of a tentative UN-facilitated roadmap for a ceasefire in Yemen, seeing continued ambiguity about Saudi Arabia's role as a mediator or party.⁵² Meanwhile, domestically AA has steadily pursued its social and religious ideological programme, as evidenced, for example, in social restrictions it has tried to impose, in how it marks anniversaries and public events, and in the summer schools it promotes for Yemeni children.⁵³

Noting all this, what changes should be recommended for efforts to bring about peace in Yemen now or in the near future? For international policy-makers and practitioners, the most important lesson to draw from the past decade of peace efforts in Yemen is that to facilitate progress on peace they must do more to engage with Sana'a and Yemen as they are, rather than as they wish or imagine them to be. In particular, those engaged in peacemaking in Yemen must change the international vision and framework that have dominated in the past decade. Some diplomats in Arab and western capitals, and officials in international organizations, are aware that the prevailing approach to peacemaking in Yemen has not worked. They should not delay making changes.

External actors (states and organizations) need to acknowledge realities about the war and the rival governments in Yemen, about what governance and peace in Yemen can realistically be, and about what is realistic for a peace process in the short and medium terms. It would help the effectiveness of mediation—whether UN-led or involving other actors—for inappropriate insistence on UNSCR 2216 to end, and indeed for the resolution to be formally superseded. To the extent that external actors can facilitate inter-Yemeni dialogue, they should do so, but without making the mistake of trying to direct a process and formulate agreements.

Changing the vision and framework does not necessitate treating the Sana'a government as the sovereign government of the entirety of Yemen, nor does it require dropping support for the aspiration of other Yemeni political forces to have more influence. Rather, it requires even-handedness in acknowledging what each government is capable of and what each controls (for better or for worse): neither should be treated as the sole government of the entirety of Yemen, or only as a militia or politicians in exile. Such an approach does not entail a significant change in position about statehood, sovereignty, borders or governance; however, it does require greater acknowledgement that hybrid and multiple arrangements in governance can exist within Yemen, as they already do, and be compatible with peace (as evidenced by the degree of peace already existing within the country). In the short and medium terms, a form of federal or confederal governance arrived at by Yemenis may be more conducive to peace than a unitary state under a single national government (of which Yemen has had little experience) or other arrangements prescribed by normative international mediation.

Changing the international vision and framework to accommodate such shifts will not end the difficulties of making and building peace in Yemen: many other factors still matter. It will no more enable AA to gain control of the whole of Yemen than the past ten years have enabled the IRG to do so. However, it will ease a significant constraint on progress and widen the space for the potential development, over the years ahead, of a Yemeni vision and Yemeni-led peace process, which are and will be what matter most for peace, whatever may happen in the near term.

For researchers on peacemaking more widely, Yemen's peace process during 2014–2024, accurately understood, is certainly a case that supports a general recommendation for more qualitative approaches in mediation research.⁵⁴ The example of persistence—over such a long period—with an unrealistic vision and an inappropriate framework also lends itself to recommendations about the gaps between mediation research and the ‘world as it is’, and the need for more reflection and evaluation in international mediation and peacemaking.⁵⁵ More than this, though, Yemen's peace process should be noted by researchers for what it reveals about the methods of international mediation. It is a case where lack of national or ‘local’ ownership and leadership of the process, and poor alignment with the context and situation, have been key factors in failure, though not its sole determinants. Shortcomings in the ownership of a peace process (who shapes and leads it), and in its alignment with local context (how well the vision and framework for a process accord with the realities of the country and conflict), are a common but under-recognized problem in peace processes. Yemen's process has been a costly example of this.

Footnotes

Author notes