Frozen Conflicts: Predicting the Sites of Future Wars

The common assumption is that every war has an end point — a discrete moment in time when the conflict is resolved, one side stands triumphant, and lasting peace takes hold. Sometimes that end takes the form of a dramatic defeat, such as the Fall of Berlin. Other times it comes from painstaking negotiations, such as those that ended the Troubles.

Other wars can simply fizzle out or be interrupted by much-bigger crises. But there is another category: the wars that don’t end, but are simply placed in the cooler — the fighting put on ice, but the deeper issues never solved. Known as frozen conflicts, they represent some of the most likely sites of future wars.

Dotted across Europe, Asia, and Africa, frozen conflicts were once thought to be manageable, hangovers from the days of decolonization or the end of the Cold War that could be contained with watchful diplomacy. But not anymore. In the past few years, multiple frozen conflicts have reignited — Nagorno-Karabakh in the Caucasus, the Western Sahara Conflict in North Africa.

Even the devastating Ukraine War exploded out of the icy conflict in Donbas. The worrying part is that there could be more to come. From Taiwan, to Kosovo, to northeast Syria, the world is riddled with frozen conflicts that could heat up again at any moment.

Defining Frozen Conflicts: Twilight Lands Without Peace Treaties

Aptly, for a phrase often used regarding contested regions, the precise meaning of the term “frozen conflict” is itself hotly debated. Some use it to mean any war that finished without an official peace treaty, even if the post-conflict status quo is now a settled fact. The Wilson Center, for example, defines frozen conflicts as places where fighting took place and has come to an end, yet no overall political solution, such as a peace treaty, has been reached.

Such a definition is also used by the Peace Research Center in Prague, and naturally covers a wide spectrum. This includes the Korean War, which stopped in 1953 but without North and South Korea ever signing a formal peace. But it also includes things like Israel and Iraq, which fought one another to a ceasefire in 1948, which Baghdad subsequently refused to sign.

For the purposes of this analysis, that is really too broad a definition. After all, a renewed Iraqi-Israeli War is unlikely to be on anyone’s bingo card for the 2020s. A much narrower definition of frozen conflicts is more useful — one focused on a specific, internal dynamic for the nations affected.

A dynamic that involves a conflict ending not just with a lack of peace, but with the ongoing existence of a piece of territory outside the central government’s control. A major, relevant example is China and Taiwan. After the Chinese Civil War ended in 1949 with a Communist victory, nationalist forces under Chiang Kai-shek fled to Taiwan, where they established a separate state.

Like North and South Korea, both Beijing and Taipei today claim the entirety of China as their own. The major difference in this case, though, is that the world’s institutions tend to agree. While both North and South Korea sit in the UN, Taiwan lost its seat to China back in the 1970s.

Today, only 13 nations recognize Taiwan’s independence. In many ways, this is a model for frozen conflicts: a situation in which a war was fought, peace found without a deal, and in which a chunk of territory remains beyond the grasp of the political center as a semi-recognized or unrecognized state — and one in which violence has the potential to re-erupt at any moment. Grouping all these conditions together is important, because it allows for likewise recognizing what is not a frozen conflict.

Northern Ireland fails the test, because — while there are still deep divisions over the issue of unification with Ireland — the Good Friday Agreement created a constitutional solution for this, in the form of a future referendum. Equally, it is important to only count those places with potential for violence. Western Sahara, for example, was a frozen conflict for decades.

But because fighting restarted in 2020, it would no longer be considered “frozen”.

Mapping Europe’s Frozen Conflicts: Kosovo, Cyprus, and Northeast Syria

With a working definition in hand, candidates around the world become identifiable. In Europe, the biggest example is Kosovo: both because of Kosovo’s semi-unrecognized nature within Serbia, but also because of Serb enclaves within Kosovo that are mostly outside Pristina’s control. Still within the EU — if not part of the European continent — the island of Cyprus has been divided since 1974, with the northern half acting as a de facto state backed by Turkey.

Right on the other side of Turkey’s borders is a far more recent frozen conflict. Northeast Syria has been ruled as an unrecognized state by Kurdish forces since the early days of the Syrian Civil War. All of these unresolved conflicts fit the narrower definition of “frozen.”

They also broadly adhere to two extra criteria laid down by the Foreign Policy Research Institute. One: the longer a conflict remains frozen, the prospect of a solution based on mutual compromise becomes more unlikely. And two: the longer the conflict remains frozen, returning to the pre-war status quo becomes progressively less likely.

But this is still just one class of such conflicts. There is a second class — the ones the term “frozen conflict” was initially identified with. A special class, all of its own: post-Soviet frozen conflicts.

It is here that some of the best-known of all can be found. Transnistria, in Moldova. South Ossetia and Abkhazia in Georgia.

Nagorno-Karabakh on the border of Armenia and Azerbaijan. At first glance, there is little to differentiate these from the ones mentioned earlier. Look a little deeper, though, and it soon becomes clear why the post-Soviet frozen conflicts are in a class all of their own.

They have all been deeply influenced by one malign actor.

Post-Soviet Frozen Conflicts and the Kremlin’s Manipulation

When the Soviet Union collapsed in 1991, it didn’t just herald the death of decades of Communist dreams. It also heralded the birth of a slew of new countries, as its constituent republics raced for the exit. Unfortunately for peace prospects, the majority of these new nations came with significant minorities.

During the heyday of the USSR, national minorities had been granted certain levels of devolution. A “full” Soviet republic like Georgia might itself contain a second-level republic like the Abkhaz ASSR, but also an autonomous oblast like South Ossetia. In each case, the autonomy granted to these lower-level regions might recognize an ethnic or religious difference, or just echo how things used to be back in Tsarist times.

But the reasons behind this complex system don’t really matter. What matters is that several of these lower-level republics and autonomous oblasts emerged from the ruins of the Soviet empire wanting their own independence. The result was a series of wars fought in the post-Soviet period.

Some of these — like the Chechen Wars — were brutally resolved. But enough wound up becoming frozen that “post-Soviet conflicts” became its whole own thing. One thing they nearly all shared: excessive interference from Russia.

The Transnistria Conflict in Moldova, the two civil wars in Georgia, and — much later — the Donbas Conflict in Ukraine all featured Kremlin machinations, often in the form of Russian boots on the ground as “peacekeepers.” Russia presented itself as a neutral actor that could help negotiate and maintain peace. Each time, though, it openly favored the breakaway regions: propping them up with economic assistance, selling them cheap oil, handing out Russian passports, and in some cases even recognizing them as independent states.

Officially, the Kremlin’s line has long been that it was doing this to protect ethnic Russians trapped as minorities in post-Soviet states. Unofficially, though, there is a much simpler explanation. The existence of these frozen conflicts long allowed Moscow to control and manipulate the countries on its borders.

The advantages to these conflicts were two-fold. One: they stopped the states engaged in them from drifting toward the West, since both the EU and NATO will not accept countries that cannot control their borders. Famously, this was what torpedoed Georgia’s chances of joining NATO.

As Condoleezza Rice later told the New York Times, the German position was: “You could not take in a country with a frozen conflict like Georgia.” The second advantage is that they kept former Soviet states firmly within Russia’s sphere of influence, unable to ignore or back away from Moscow’s demands, or forge their own destinies.

Moldova, Ukraine, and the Kozak-Minsk Playbook

Moldova offers a clear illustration. Since a 1992 war, the tiny former Soviet republic has been cleaved in two — with a sliver of land in its far east under the control of a self-proclaimed state known as Transnistria, a state guarded by roughly 2,000 Russian peacekeepers. Because the conflict is frozen, Moldova has been unable to integrate further with the West.

At the same time, Moscow has been able to use the existence of its proxies on Moldovan territory to manipulate Chișinău. This influence has been so useful that the Kremlin tried to formalize it. In 2003, Russia pushed the Kozak Memorandum as something that would achieve lasting peace between Moldova and Transnistria.

The problem: the Memorandum envisaged turning Moldova into a federal state in which each constituent part would be empowered to veto all major decisions. In practice, this would have meant Putin could phone up his allies in Transnistria and ask them to kill any Moldovan legislation he didn’t like. The frozen conflict would have ended, but at the cost of Moldova’s sovereignty.

The exact same logic is at play in the Minsk 2 Agreement to end the 2014 conflict in Ukraine’s Donbas. Proposed when Ukrainian forces were getting hammered by Russian-backed separatists, Minsk 2 mostly dealt with ending the violence and holding elections. However, it also contained an infamous paragraph: Article 11.

Under Article 11, the Russia-backed Donetsk and Luhansk People’s Republics would have been free to pursue their own relations with Moscow, independent of Ukraine’s government. They would have been able to create their own police force and choose judges and prosecutors. Even worse were the May 2015 proposals Russia pushed, but which Ukraine never signed nor ratified.

These included giving the statelets full control of the Russia-Ukraine border; their own state budgets; the right to pursue their own foreign policy; and the power to hold referendums or declare states of emergency. Ukraine would have had to write a neutrality clause into its constitution — “neutrality” in this case being Russian for “doing whatever the Kremlin wanted.” Just like the Kozak Memorandum, Minsk 2 was never fully implemented, for the good reason that it would have turned Ukraine into a mere Russian puppet.

But its stipulations — negotiated with the help of France and Germany — offer a clear insight into Vladimir Putin’s reasons for indulging these frozen conflicts.

Beyond Russia: Taiwan, Cyprus, and the Kurdish AANES

Kremlin manipulation can likewise be seen in the Nagorno-Karabakh Conflict — a once-frozen war that differed from its post-Soviet counterparts by not involving a Russian minority in its breakaway state. Instead, ethnic Armenians fought for independence from Azerbaijan in the mid-1990s, leading to a stalemate. But even this could still be manipulated by Moscow.

As Armenia’s military guarantor, Russia used Nagorno-Karabakh to keep Yerevan from drifting toward the West. In 2012, worry of renewed war with Azerbaijan made Armenia ditch a planned association agreement with the EU. Still, not all frozen conflicts feature Russian manipulation.

The best-known frozen conflict on the planet is probably the unresolved civil war between China and Taiwan. Utterly unrelated to Russia or the USSR, this long-frozen war nonetheless features an extremely similar dynamic. There is a mostly unrecognized state within the central government’s internationally accepted borders that is backed by outside actors: Taiwan is backed by the USA, but also much of Europe.

At the same time, the deeper issues remain unresolved, and the potential for violence is there, bubbling away. Xi Jinping has made a central goal of his rule the reconquest of the breakaway island. The history of North Cyprus is exceptionally complex.

In 1963, inter-ethnic violence erupted on the island between Greek and Turkish Cypriots, violence that eventually exploded into the 1974 Turkish invasion. What followed was a dramatic, four-week war that killed thousands and left Cyprus bitterly divided. In the aftermath, the Turkish Cypriots in the north declared independence, a move only recognized by Turkey.

Even now, decades later, Ankara still maintains something in the region of 35,000 troops there. The Turkish government uses this to its geopolitical advantage. Like Moscow, Ankara wants to keep Cyprus from drifting too close to the West.

While the island was able to join the EU in 2004, it is not within NATO. Ankara wants access to offshore hydrocarbon resources, as well as to use Cyprus’s coast to project its naval power. By this point, it becomes clear that the existence of a powerful outside actor is another necessary condition for these frozen conflicts: Russia in most post-Soviet spaces; the USA as a security guarantee for Taiwan; and Turkey on the island of Cyprus.

Turkey is in fact connected to multiple conflicts. In Nagorno-Karabakh, the main tensions are between Armenia and Azerbaijan, but the unfinished war is also a proxy fight between three local frenemies: Turkey on the Azeri side, and Russia and Iran on Armenia’s. Ankara is also a key player in one of the newest such conflicts: northeast Syria.

During the Syrian Civil War, the Assad regime effectively lost control of the country, leading to the rise of multiple statelets, warlords, and — most infamously — the ISIS caliphate. Up in the northeast, something more enduring emerged. The Kurdish Autonomous Administration of North-East Syria (AANES) is an unrecognized state of about 5 million that controls many of Syria’s oil and gas fields.

The People’s Protection Units (YPG) were instrumental in the defeat of ISIS, and their soldiers still guard tens of thousands of captured militants. The reason AANES continues to exist is due to a hyper-complex swirl of relations between three major powers: the USA, which helped the Kurds defeat ISIS and now feels a duty of care toward them; Turkey, which considers the AANES to be linked to Kurdish militants in Turkey and wants to destroy it; and Russia, which backs Assad and is against the AANES.

How Frozen Conflicts Reignite: Nagorno-Karabakh and Western Sahara

For most outsiders, the violence seemed to blow up out of nowhere. After decades of tense peace — bar a four-day dust-up in 2016 — the contested Nagorno-Karabakh region exploded in the fall of 2020. As most of the world was glued to news of the ongoing pandemic, Armenia and Azerbaijan fought a brutal, six-week war over the ethnic Armenian enclave within Azeri territory.

Perhaps 7,000 were killed. Tens of thousands of civilians were driven from their homes. By the time the dust settled, Azerbaijan had successfully conquered swathes of territory once de facto controlled by Armenians.

Yet this was not the end. Two years later, in the fall of 2022, Baku launched a fresh attack — one that lasted a mere two days but still managed to kill around 300 people and see yet more territory in Nagorno-Karabakh fall to Azeri forces. The conflict had frozen way back in 1994.

Aside from those four days in 2016, it had remained that way ever since. So what happened? What caused an effective 26-year ceasefire to suddenly collapse into horrendous violence?

The answer is a terrifying study in how quickly outside factors can reignite these conflicts. Back when the 1994 ceasefire was declared, Armenia was the effective winner. Backed by Russia, Yerevan had been able to bring firepower to bear that Azerbaijan just couldn’t match.

In the quarter century since, though, regional politics had undergone a seismic shift. Massive investment in oil and gas extraction had boosted Azerbaijan’s state budget — a budget it then spent buying advanced military equipment from its powerful backer, Turkey. And while Armenia was still technically backed by Russia, Vladimir Putin had been growing personally closer to both Azerbaijan’s Ilham Aliyev and Turkey’s President Erdoğan.

In particular, Russia had closed a major deal in 2019 to supply Turkey with S-400 missile defense systems — a blow to NATO that Putin did not want to jeopardize. By 2020, Baku was clearly able to sense it had the advantage. In the opening hours, Turkish-bought TB2 drones destroyed Nagorno-Karabakh’s air defenses, turning Armenia’s tanks into sitting ducks.

Although Moscow would eventually impose a ceasefire, it would not be until its supposed ally Armenia had lost thousands of square kilometers of territory. The two-day follow-up in 2022 had an even clearer cause. With Russia badly bogged down in Ukraine, the Azeris gambled that Russian peacekeepers installed after 2020 would not dare interfere.

What followed was a wholly opportunistic land grab that took place at the exact moment the Kremlin was reeling from Ukraine’s Kharkiv Counteroffensive. In short, a shift caused one side to recalculate its risk-reward ratio and decide conditions were now favorable. Actors in these conflicts do not respond only to opportunities and positive changes.

They might also be reacting to a negative change — or even an absence. Such was the case with the Western Sahara Conflict, which resumed in 2020 after a 29-year ceasefire. During decolonization, Spain agreed to the creation of two independent countries: Morocco, and the Sahrawi Arab Democratic Republic.

But Morocco instead claimed all territory as its own. This sparked a war that dragged on until 1991, when a ceasefire was agreed on the basis of a future referendum over Western Sahara’s sovereignty, as well as the creation of a UN buffer zone. What caused fighting to re-erupt in 2020 involved road blockades, violations of the UN buffer zone, and grievances related to Morocco never holding the agreed referendum.

But it was at least in part fueled by the sudden resignation of the UN envoy to the region and the subsequent months-long suspension of negotiations. With an acceptable path to peace closed off, representatives of the Sahrawi Arab Democratic Republic felt they had no choice but to take action.

The Dangerous Instability of Unfinished Wars

The northeast Syria situation is a clear example of how complex these frozen conflicts can be — how they are sustained not just by facts on the ground but by a delicate dance between powerful outside forces, a dance that makes them inherently unstable. There is a reason the subject carries the subtitle “predicting the sites of future wars.” It is the horrible ease with which the slightest change in either local, regional, or big-power relations could heat any of these conflicts back up.

In this sense, “frozen conflicts” is probably the wrong name for them. More apt might be something like “Jenga conflicts,” because of the way the tiniest shift could collapse things in an instant. The situation between Kosovo and Serbia illustrates this well.

There, Kosovo is the breakaway, semi-recognized state backed by powerful outside actors: the US and NATO. But it also contains its own breakaway region: the Serbian-majority enclaves in the north, which reject Pristina’s authority and run themselves as semi-independent statelets. They are backed by Serbia, which is pushing a peace settlement known as the Community of Serb Municipalities, which would give these enclaves massive amounts of autonomy and the power to appoint their own police forces and judiciary.

Serbia itself is backed by Moscow — a set of nesting Matryoshka dolls of contested sovereignty. Like Taiwan, like Cyprus, like AANES, or Transnistria, Abkhazia, or South Ossetia, Kosovo is a complex place — one where the war has both been over for years, yet simultaneously never ended. A place where conflict could potentially break out again with little warning.

That is the defining feature of these frozen conflicts: their ability to quickly heat back up, to quickly go from uneasy peace to brutal warfare. Over in Asia, Xi Jinping might decide the balance of military power has tipped far enough in his favor to move on Taiwan. In northeast Syria, a change of relations with the US could leave Turkey feeling it has carte blanche to move against the AANES.

A total Russian defeat in Ukraine might empower Georgia to reclaim Abkhazia and South Ossetia. Conversely, Putin might try to stop Moldova’s urgent moves to join NATO by reigniting the Transnistria War. None of these are firm predictions about how the future may play out.

What they are, though, are warnings — warnings on the dangerous nature of these unfinished conflicts, on the ways they can destabilize not just the states they take place within, but also, potentially, the entire world. It could be that none of these frozen conflicts ever reignite. It could be that most, or even all of them do.

All that is certain is that, when it comes to predicting the sites of future wars, it is these frozen conflicts that may hold the greatest potential for bloodshed.

Frequently Asked Questions

What precisely is a frozen conflict, and what distinguishes it from other unresolved wars?

For the purposes of this analysis, a frozen conflict is a war that ended without a peace treaty and left a piece of territory outside the central government’s control, typically backed by an outside power, with potential for violence to reignite at any moment. This narrower definition excludes cases like a future Iraqi-Israeli war — highly unlikely — while capturing genuinely unstable situations such as Taiwan, Kosovo, and Transnistria. The key additional criteria from the Foreign Policy Research Institute are that the longer a conflict stays frozen, the less likely mutual compromise becomes, and the less likely a return to the pre-war status quo is.

How has Russia used frozen conflicts to manipulate neighboring states?

Russia has systematically used frozen conflicts to prevent former Soviet states from joining the EU or NATO, since both institutions refuse members that cannot control their own borders. By propping up breakaway regions in Moldova, Georgia, and Ukraine with economic aid, cheap oil, Russian passports, and in some cases formal recognition, Moscow kept those countries dependent on and unable to ignore the Kremlin. Both the 2003 Kozak Memorandum — which would have given Transnistrian proxies a veto over Moldovan legislation — and the Minsk 2 Agreement’s Article 11 illustrate how Russia tried to formalize this leverage into permanent political control.

What caused the 26-year ceasefire in Nagorno-Karabakh to collapse in 2020?

The ceasefire broke down because the underlying balance of power had shifted dramatically. Azerbaijan’s oil wealth funded major military modernization, while Turkey supplied Baku with TB2 drones that destroyed Nagorno-Karabakh’s air defenses in the opening hours of the 2020 war, turning Armenian tanks into sitting ducks. Armenia’s nominal Russian backing proved unreliable: Putin had grown closer to both Aliyev and Erdoğan and had just sold Turkey the S-400 system — a deal he did not want to jeopardize. The 2022 follow-up attack was timed even more deliberately, launching at the exact moment Russia was reeling from Ukraine’s Kharkiv Counteroffensive and its peacekeepers dared not interfere.

Why is Turkey simultaneously involved in multiple frozen conflicts?

Turkey is a key outside actor in at least three such conflicts at once: it backs Azerbaijan against Armenia in Nagorno-Karabakh, maintains roughly 35,000 troops in North Cyprus to prevent the island from drifting fully toward NATO and to access offshore hydrocarbon resources, and seeks to destroy the Kurdish AANES in northeast Syria, which it views as linked to militant Kurdish groups operating inside Turkey. This web of entanglements means that a shift in Turkish policy or U.S.-Turkey relations could simultaneously destabilize multiple frozen conflicts across a broad geographic arc.

What would it take for any of these frozen conflicts to reignite?

History shows that a relatively small shift in the risk-reward calculus of one side is enough. A change can be positive — like Azerbaijan sensing military advantage in 2020 — or negative, like the Western Sahara conflict resuming in 2020 after the UN envoy resigned and negotiations collapsed, leaving no acceptable path to peace. In other cases, a distant event creates an opening: Azerbaijan exploited Russia’s distraction during the Kharkiv Counteroffensive in 2022. For Taiwan, Kosovo, or the Kurdish AANES, a change in U.S. commitment, a shift in China’s assessment of military advantage, or a recalibration of Turkish or Serbian ambitions could each serve as the trigger that turns an uneasy peace into open war.

Sources

1. https://warsawinstitute.org/post-soviet-frozen-conflicts-challenge-european-security/
2. https://www.prcprague.cz/fcdataset
3. https://www.chathamhouse.org/2020/05/minsk-conundrum-western-policy-and-russias-war-eastern-ukraine-0/minsk-2-agreement
4. https://www.clingendael.org/publication/turkeys-interventions-its-near-abroad-northern-cyprus
5. https://www.brusselstimes.com/486194/kurdish-north-east-syria-frozen-conflict-with-the-eu-in-the-backseat
6. https://www.cfr.org/global-conflict-tracker/conflict/nagorno-karabakh-conflict
7. https://www.cirsd.org/en/young-contributors/russias-approach-to-frozen-conflicts-studying-the-past-to-prevent-the-future
8. https://www.aljazeera.com/news/2020/11/13/morocco-launches-operation-in-western-sahara-border-zone