Supporters of the Houthi movement shout slogans during a celebration on Sept. 26 just days after seizing much of the Yemeni capital of Sanaa. (Yahya Arhab/EPA)

This week’s events in Yemen have been mischaracterized in a variety of ways: A sectarian conflict between Shiite and Sunni groups; a second revolution that finally removes the pre-2011 actors from power; a counter-revolution backed by former president Ali Abdullah Saleh’s party and his tribal affiliates; and a self-staged coup condoned by President Abed Rabbo Mansour Hadi to weaken his political opponents. Given the rapidly evolving events, the malleable and shifting alliances, and the profusion of backroom deals among the main political actors, such a menagerie of competing explanations – each of which contains a nugget of truth – is not surprising. Nevertheless, a longer look at how political actors in Yemen balance against each other, and at how a shift in such balance of power between groups that is not reflected in the distribution of power in the government, provides a better explanation for the crisis. It also provides lessons about how to avoid similar escalation in the future.

The best way to conceptualize Yemeni politics is through the concept of anarchy. By this, I do not mean disorder and chaos – although the visual imagery it evokes is not entirely unwarranted – but rather the specific meaning of the term in international relations scholarship: The absence of a sovereign capable of enforcing rules of engagement. Under anarchy, multiple sovereigns inhabit the same political space, and there are no assurances that agreements will be kept because there is no higher authority to enforce compliance. As such, all actors must be alert to the possibility for other actors to behave aggressively or to renege on agreements, which makes alliances between them unstable in the long run. It is this concept of anarchy that most distinguishes international relations from domestic politics, and it is by transcending anarchy that the centralized, hierarchical political systems present in functioning states are founded.

From the unification of Yemen until the uprisings of 2011, political power was generally centralized. This was achieved not by an absence of opposition – opposition to the government has always existed in Yemen, even in its most extreme, armed, form – but through the relative stability of the balance of power between opposed forces. Saleh and his allies were adept at temporarily assuaging grievances through patronage networks, pitting various groups against one-another and, when these failed, deploying brute force.

Protests in 2011 brought these marginalized groups to the center. Prior to that, the Houthi movement, which began as Zaydi Shiite opposition to the central government’s repression of religious minorities and willful underdevelopment of their areas of the country, had little support outside of its traditional power center in northern Yemen. From 2004 to 2010, the Houthis fought six wars with the central government, during which time they developed a hardened and effective military wing.

In 2011, the Houthis allied briefly with Gen. Ali Mohsen al-Ahmar, one of the most powerful men in Yemen and the chief architect of the wars against Houthis, as well as with the al-Islah party and other revolutionary groups protesting Saleh’s regime. These alliances were short-lived, however. The transitional government that was established with Gulf Cooperation Council (GCC) brokerage in November 2011, which gave Saleh immunity from prosecution, was split equally between Saleh’s General Congress Party (GCP) and the former opposition bloc, the Joint Meeting Parties (JMP), of which Islah is the prominent member, while leaving out the Houthis. The Houthis rejected the transitional government, as well as Saleh’s immunity, feeling that none of their grievances had been addressed.

Therefore, not only did the transitional government not reflect the shifting balance of power in the country now that both the Houthis and Southern separatists represented large constituencies and militias, all formal political positions remained in the hands of the existing elites. The revolution left Yemen incredibly fractured: The military was split into various blocs (with part of the army loyal to Mohsen, who turned against Saleh during the 2011 protests, part of it under the authority of Saleh’s son, and another nominally under Hadi); the Houthis had complete control over vast areas of the North; and the state had virtually no control over the South. Even this is a simplified picture, which does not take into account areas under the control of Salafi groups or tribal associations. The revolution, therefore, brought about the devolution of Yemen from a relatively centralized (and highly corrupt and repressive) state into various virtual states-within-states.

The main purpose of the National Dialogue Conference, which ended in January, was the establishment of the framework for the transition, electoral process, and division of the country into federal states, though it fell short in a number of ways. It did not address the vastly important issue of equitable powersharing, failed to satisfy the demands of the Houthis and the Southern separatists concerning federal borders, and had enormous flaws in its implementation. While the Houthis were praised for their pragmatic stance during the dialogue, they were simultaneously engaged in violent battles against various affiliates (such as tribes and Salafi groups) of the Islah, Ahmar tribal group, and Mohsen bloc.

The Houthis have shown profound pragmatism and strategic planning over the past three years. In addition to their consecutive military victories in the North, they tapped into popular resentment against the continued rule of the old elites, the removal of fuel subsidies this summer, and the slow (practically nonexistent) pace of reform. Of these, the reinstatement of the fuel subsidies became a rallying point with groups unaffiliated with the Houthis. While the removal of the subsidies was a financial necessity that would have had the added benefit of eliminating a large source of income from the patronage networks of elites, it was done abruptly and without consultation. The cost to Yemenis was only matched by the political cost of such an unpopular move.

The Houthis, therefore, claim the support of their own Zaydi constituency and their allies in the Southern movement, as well as vast numbers of Yemenis who view them as a real opposition to the elites that is untainted by corruption. This by no means implies universal support, because in addition to the armed opposition to them, many Yemenis are worried that the Houthis will force a Zaydi rule and be as repressive toward the rest of Yemen as they have been reported to be in areas already under their control. However, this type of wider support that goes beyond their Zaydi stronghold is sufficient reason to reject simplistic binary explanations of the conflict as a Shiite versus Sunni sectarianism.

From January until late August, the Houthis marched steadily toward the Yemeni capital of Sanaa after winning battles against Salafi affiliates of Islah, the Ahmar tribes (Islah’s strongest military support) in their traditional strongholds, and battalions loyal to Mohsen. Hadi made what seemed at the time to be the strategically smart choice not to intervene in the conflict: Not only did he lack the military power to stop it, but such intervention could easily have exacerbated the situation into a full-blown civil war. He also benefited from this conflict insofar as each party weakened the other and so, as Saleh had done before him, he let the conflicts continue. Rhetorically, however, there was much condemnation of the Houthi threat and the danger of civil war, both domestically and internationally.

In an ironic twist of events, the Houthis also currently enjoy the support of tribes loyal to Saleh and his party. While neither party would accept a formal association between them, it serves them both to fight their common enemies. This is, of course, a complete reversal from 2011 to 2012. But if one thing is predictable in Yemen, it is that alliances change, and that they will do so unpredictably. Alliances depend on the relative balance of power between actors and, as this balance has shifted, almost all of these main groups have, however briefly at times, concurrently shifted from open aggression to alliances of convenience.

Any analysis of Yemen would be incomplete without a consideration of regional powers, as well, however. The Houthis have been accused of being pawns of Iran. Their traditional slogan (“Death to America! Death to Israel!”) and military tactics resemble those of Iran’s other ally, Hezbollah. Meanwhile, Saudi Arabia has tried repeatedly to squash them through aerial attacks, and has for years supported various evangelical Salafi groups as well as Islah (which is composed of the Muslim Brotherhood, Salafi groups and businessmen) in their fights against the Houthis. The extent of Iranian involvement is debated, but the Saudi animosity toward the Houthis is well-known.

Yet, events in the rest of the Middle East have shifted Saudi preferences in Yemen, too. With the defeat of the Muslim Brotherhood in Egypt, Saudi Arabia took the opportunity to weaken the Muslim Brotherhood in Yemen by slowly withdrawing support from Islah. Similarly, events in Syria and Iraq have raised fears of extremism in the region and, given that Yemen is already home to al-Qaeda’s branch in the Gulf, a stable Yemen is seen as a political necessity for regional stability. All parties must have recognized that a stable Yemen is entirely impossible as long as too many power centers compete with each other. Hence, given the Houthis’ growing power over the last few months, it made strategic sense to throw their support behind the winning team. The GCC gave repeated warnings against the Houthi advance toward Sanaa just a few weeks ago, and its endorsement of the recent events would seem incoherent if seen outside of the context of regional politics. Iran, Saudi Arabia and the United Nations have welcomed these developments (although the latter more cautiously), which would have seemed unfathomable just a few weeks ago.

The rapidity with which Houthis overtook Sanaa must be seen at least partially in this context. Hadi and the army offered no resistance to the advancing Houthi militias, who only encountered resistance from Mohsen’s forces. In fact, the Ministry of Interior quickly issued a statement of cooperation. In this sense, these events do seem like a self-staged coup, even if only by remaining on the sidelines: Hadi and the Gulf states hung Islah out to dry through their neutrality, which had the indirect (but possibly intentional) effect of empowering the Houthis.

These sorts of rapidly shifting fortunes are precisely what should act as a warning in the phases ahead. The Houthis and the government signed a deal this week that will give both the Houthis and the Southern separatists ministerial posts and advisory roles in the presidency. In their infinite pragmatism, this past week the Houthis have even ignored the “Death to America/Israel” in their slogan and moved toward an inclusive, populist, social justice narrative and domestically focused calls for unity. Their leader, Abdel Malik al-Houthi, gave a speech upon taking control of government offices in which he explicitly focused on inclusivity and democracy as the goals of the movement and as their vision for the future of Yemen. He also, however, refused to sign an annex to the deal that would call for disarmament and withdrawal of Houthi troops.

What does this mean for the future of Yemen? This will depend on how well the lessons of the past are learned. This could provide a fresh start for Yemen, with some of the most important political figures so severely weakened. It could, also, give Saleh and the GPC a greater role again, which would be a true counter-revolutionary move. And, worse, the Houthis could betray their rhetoric and give themselves an outsized slice of the pie. Given that Islah, Mohsen and al-Qaeda have already declared war on the Houthis (possibly with another backroom deal among them), this would be greatly destabilizing.

Yemen remains fractured between various centers of power and although Islah, the Ahmars and Mohsen are weakened, they are not defeated. The only way out of this situation is the creation of a government that reflects the real balance of power and is inclusive in practice. That a lot of this will depend on whether or not the Houthis decide to share some of the power they already have is worrisome, not only because of their behavior in their own region, but because few things in politics resemble the laws of physics: Actors not willingly giving up power is, however, one of them.

Silvana Toska is a PhD candidate in political science at Cornell University, with a regional focus on the Middle East. She is currently writing her dissertation on the causes of revolutionary waves.