The Role of Security Guarantees

How will Ukraine fit into to the growing web of Western alliances? And are they a force for good?

The intense negotiations at the Vilnius summit about how and when Ukraine will join NATO need to be understood in the context of the ongoing war with Russia, and what this signals about the West’s readiness to provide for its long-term security. But it also offers an opportunity to reflect on broader changes underway in the character of alliances and their international roles.

In the past decade, Western military alliances – those which include a security guarantee – have undergone a significant transformation. They have enlarged and deepened. They also have been supplemented by more informal defence and security links. Non-Western alliances are changing too: China-Russia ties were reinforced prior to the war, al though overall Russia’s system of alliances is in crisis. What does this mean for global stability? Will it make the world more stable, or more dangerous? And how should we look at security guarantees for Ukraine in that context?

The enlargement and deepening of Western alliances

Established institutions – especially those with entrenched bureaucracies – tend to grow rather than decay. But they will do so only if the there are continuing rationales for their existence. This process has taken place for the European Union (EU) and the North Atlantic Treaty Organization (NATO).

The EU can now be described as a military alliance. The Treaty of Lisbon commits EU members to collective defense through its article 42.7. The Union has continued to enlarge with the accession of Croatia (2013) and despite Brexit is likely to welcome new members in the coming years. NATO too has enlarged with the accession of Montenegro (2017), North Macedonia (2020), and Finland (2023), while Sweden is expected to join very soon. Members have updated their security outlook including by making it clear that forms of non-traditional aggression in cyberspace (2014) as well as “to, from or within outer space” (2021) could qualify as armed attack and trigger Article 5.

Outside of Europe, the web of Western alliances has also developed. During the 2010s, Washington instituted an Extended Deterrence Dialogue with Japan and an Extended Deterrence Strategy and Consultation Group with Korea. In 2014, the US bolstered its defense commitment to the Philippines and made it clear that an attack against Filipino forces in the South China Sea would be covered by the 1951 treaty. Washington is also reinforcing its defense cooperation with Australia, and both countries now agree that their alliance is also valid for defending against cyber-attacks. The Biden administration has confirmed that it would defend Taiwan against Beijing, rhetorically putting the island on a par with countries protected by treaty-based obligations. Even Saudi Arabia is showing interest in having a formal alliance with Washington.

In addition to formal agreements, a slew of defense cooperation groupings have also emerged in recent decades. The United States has considerably expanded the list of its major non-NATO allies, which number 19 today against five only in the late 1980s. Nowhere is this expansion of informal security groupings clearer than in the Indo-Pacific. The AUKUS security partnership set up in 2021 brings together Canberra, London and Washington. US-New Zealand defense cooperation is now being rejuvenated. India has been granted a special status of “Major Defence Partner”. The 1971 Five Power Defense Arrangements are being given a new life, due in particular to British and Australian interest.

Why did alliances endure?

It was not a given that Western alliances would survive and thrive. President Donald Trump wondered whether Washington should defend the “tiny” and “aggressive” Montenegro. President Macron wondered whether NATO was in danger of “brain death.” Spats between Turkey and its allies have led many to wonder whether Ankara would (or should) remain a member of NATO. US-Israel and US-Saudi relations – not formal alliances – have turned sour due to domestic political evolutions

The endurance of alliances can be partly explained by inertia, vested interests, and sunk costs. But there is more to it. Modern alliances are defensive in nature and are therefore more resilient than the aggression-oriented pacts of the distant past. Moreover, they are generic as opposed to threat-specific: they apply to any aggression. This has rendered them capable of confronting the re-emergence of perceived threats, or be the vehicle for dealing with new ones.

The growth of the Western system of military alliances and partnerships is mostly a consequence of the radicalization of Russian and Chinese policies (as well as North Korean and Iranian ones). To use political science jargon, there is more “balancing” going on vis-à-vis Moscow and Beijing, and more “bandwagoning” with the United States.

Then again, the US system is “unique in human history.” A scholar calculated that US alliances covered some 25% of the world’s population and 75% of its GDP. It is institutionalized like no other. NATO is not only the biggest formal alliance in the world but also one of only two (with RoK/US Combined Forces Korea) that has a permanent military command. It has proven flexible enough to lead major operations with non-members. Unlike ad hoc coalitions, NATO allows for collective decision-making, tested procedures, interoperability, and the use of common assets. US-based alliances tend to last twice as long as non-US-based ones: the main treaty-based US alliances forged after 1945 are historical outliers in terms of their duration.

The United States stands out as a security guarantor due to the combination of its helpful location, its democratic nature and it unparalleled power. US-led formal alliances are also unique in that they involve both interests and ideals. Today, the US administration is expanding and consolidating these links seeking to “build a latticework of alliances and partnerships globally that are fit for purpose for the 21st century.”

More alliances, more wars?

Is this thickening web of alliances good or bad for global stability? Does it decrease or increase the risk of war? I believe the former. 

Deterrence does seem to work. Russia and China have never openly committed armed aggression against territories clearly covered by Western security guarantees; by contrast, they have invaded or encroached upon non-protected countries or disputed territories.

As stated, today’s alliances are defensive in nature and there is solid academic evidence that such alliances provide stability and predictability. They are less likely in themselves to produce a pushback from other countries, especially when they adopt unilateral confidence-building measures aimed at reassuring a potential adversary that they have no aggressive designs. They often operate by consensus.

Almost all defense commitments have significant caveats. They do not compel allies to use military force, and where applicable they mention national restrictions such as “the constitutional provisions and processes” (US-Japan) or “the specific character of the security and defense policy of certain member States” (TEU). They also apply only to a defined area. NATO’s Article 5 applies to “Europe and North America.” This does not include US territories in the Pacific. The US-Japan treaty covers “the territories under the administration of Japan.” The US-South Korea and US-Philippines ones are drafted along the same lines.

Most of them contain sufficient ambiguity so that the protected party cannot assume that its guarantor would automatically use force to defend it. The most significant caveat is the notion of armed attack, the casus fœderis of defense alliances. The International Court of Justice (ICJ) considers that it involves a relatively large scale attack, of sufficient gravity, and with a substantial effect.

Another uncertainty relates to the exact definition of protected territories, the centerpiece of security guarantees. This is increasingly a problem at sea. Since the UN Convention on the Law of the Seas entered into force in 1994, many States have made irreconcilable claims regarding their territorial waters as well as Economic Exclusion Zones. In addition, Beijing has had an active policy of land reclamation and occupation in the South China Sea aimed at creating “facts on the water” to alter the legal status of rocks, islets and reefs it controls.

Finally, there are some lingering uncertainties inherited from the Cold War. The US commitment to its South-East Asia Treaty allies only applies to “Communist aggression.” Would contemporary China appear qualified as such by all signatories?

All this suggests that allies do not get embroiled by mere virtue of a text or declaration: going to war would remain a political decision.

Entrapment concerns

Protected parties may fear abandonment, but security guarantors fear entrapment.

Alliances may embolden member states to initiate disputes against others and be more likely to aggravate dispute initiation against member states. Recall that President Trump’s comments about Montenegro were precisely about the fear that an ally could drag the United States into “World War Three.” In particular, “revisionist countries holding unconditional deterrent agreements are more likely to initiate conflict than if they had not been given an alliance or had been given a conditional deterrent alliance instead.” But there is little evidence that contemporary allies have shown aggressiveness in their neighborhood without fear of retaliation because they felt protected. True, weaker parties can mistakenly convey to their publics that contested territories would be protected. The public debates surrounding the Russia-Armenia defense accord of 2010, or the France-Greece strategic partnership of 2021 are cases in point. But in both cases, clarifications were given by the protector.

A second issue is entrapment through reputation concerns, often raised as a reason to intervene in defense of an ally or a partner. But an analysis of US commitments supports the idea that Washington maintained its freedom of action when deciding whether or not to intervene: it found only five examples of ostensible entanglement since 1945. And even in such cases, the United States had many interests at stakes, not only those related with reputation. By contrast, Washington did not support the French in Diên Biên Phu and undermined both the French and the British in Suez. Allies in fact often play the role of a brake on escalation. “In most conflicts, only a few allies were directly threatened and demanded U.S. intervention. (..) Most allies (..) urged restraint because they worried their security would suffer if the U.S. drained its strength in a peripheral region or escalated a faraway conflict into a global war.”

NATO has been cautious in invoking Article 5: it did not do so in 2007 when massive Distributed Denial of Service attacks of Russian origin affected Estonia, or when a Turkish aircraft was downed by Syria over international waters in 2012. Seoul and Washington did not over-react when North Korea attacked a South Korean ship and shelled islands in 2010. Iranian attacks on Saudi territory over the past decade elicited fairly measured US reactions. If anything, the United States occasionally appears hesitant in upholding its own red lines.

In fact, modern defense commitments are vaguer than they were in the past in terms of the anticipated allied response. This contributes to the freedom of action of guarantors.

For all these reasons, even those countries enjoying a US security guarantee could not assume that its implementation would be automatic. This will always be a matter of political judgment. Moreover, there is of course the possibility that the guarantee will break  down even though the circumstances for which it was designed – a clear-cut armed attack – are present. Defensive alliances are not always upheld: an academic analysis finds that they are honored only 41% of the time. What if a US president refused to defend a smaller NATO member against regional aggression?

The occasional suggestion of a possible 1914-like sequence of events – in reference to how the European alliance system is said to have facilitated the march to general war by chain-ganging – seems off. Likewise with the comparisons with the Cold war.

In any case the traditional narrative of World War One as an unfortunate, uncontrollable chain of events is now challenged. In 1914, “the war began, not as a result of chain-ganging, but because of coordinated aggression by Germany and Austria-Hungary. The latest historical research on the origins of World War I is inconsistent with the chain-ganging hypothesis.”

An asymmetrical situation

Also, the early 21^st century network of alliances and partnerships is profoundly asymmetrical.

Washington has five dozen formal allies, Russia has five, and its alliance system appears increasingly fragile. To be sure, its defence relationship with Belarus, which has a union treaty with Moscow, is getting stronger. But three members of the Collective Security Treaty Organization (CSTO), an organization formed with former Soviet Republics, left in 1999 and more importantly, after its invasion of Ukraine Moscow appears unwilling or unable to support its allies. This was the case when Armenia requested Russian support in September 2022 as its internationally-recognized territory was attacked, despite the security guarantee provided by both the CSTO and a bilateral pact – something which led Yerevan to refuse signing the summit’s communiqué. The CSTO also failed to bring back order when Tajikistan and Kirghizstan clashed the same year. 

In 2021, China renewed its 1961 Treaty on Friendship, Cooperation and Mutual Assistance with North Korea, but it remains its sole formal defense commitment. Despite the closeness between Pakistan and China, Islamabad insists that it does not want to be part of any bloc.

Russia and China started developing a tighter defense partnership after signing their Friendship and Cooperation Treaty in 2001, renewed in 2021. Their bilateral has in fact turned into a near-alliance. In recent years Moscow has started to use the words “alliance” or “allied” to qualify its relationship with Beijing. However, the war in Ukraine has shown that there are clear limits to their so-called “no-limit” partnership and that they would not commit to each other’s defense.

The thickening web of contemporary Western alliances is likely to be more stabilizing than destabilizing. This network of alliances is more a safety belt than the imagined conveyor belt nightmare of 1914, and may produce a different kind of stability than the bloc-to-bloc one of the Cold war.

If the analysis above is correct, the West should not fear giving security guarantees to Ukraine.

What security guarantees for Ukraine?

While “guarantees” is a stronger word than “assurances”, the two terms are sometimes confused or used interchangeably, partly because their legal value varies considerably from one type and context to another. They are commitments to do (positive) or not do (negative) something in the realm of security and defense. All security guarantees are meant to have a reassurance value.

Security assurances are normally associated with nuclear non-proliferation. Negative security assurances (“garanties” in French) are commitments given by Nuclear Weapons States (NWS) to refrain from using nuclear weapons against Non-Nuclear Weapons States (NNWS) which are parties to the Nuclear Nonproliferation Treaty (NPT). Positive ones are commitments to assist such countries should they become victims of such attacks, when the aggression should be seized by the UN Security Council. Some of them are treaty-based (those contained in protocols to treaties establishing nuclear-weapon-free-zones).   

Security guarantees – the main focus of this post – are commitments to support a country (positive), or to refrain from taking military actions (negative) against another. Positive security guarantees can take various forms, from treaty obligations such as NATO’s Article 5 or bilateral arrangements to presidential statements such as “we will help country X… to defend itself”. Negative security guarantees include treaty-based obligations or mere political commitments such as NATO’s commitments made to Russia (no deployment of nuclear weapons or permanent presence of significant combat forces in Central Europe) after the break-up of the Soviet union and the end of the Cold War.  

The debate about Ukraine often confuses these different types of commitments. This is understandable given that the invasion of the country since 2014 shed light on the Budapest memorandums, signed in 1994. These three identical documents each included several different kinds of assurances. In exchange for returning Soviet nuclear weapons to Russia, Ukraine became a NNWS in the sense of the NPT. Three of the five NWS – the US, the UK and Russia – signed the memorandums. China and France committed themselves separately.

The five NWS (re-)committed themselves to “respect the independence and sovereignty and the existing borders of Ukraine” and to “refrain from the threat or use of force against the territorial integrity or political independence of Ukraine” (negative assurances). 

In addition, they reaffirmed their NPT-related negative (no use of nuclear weapons against a NNWS) and positive (seek immediate UN Security Council action should Ukraine “become a victim of an act of aggression or an object of a threat of aggression in which nuclear weapons are used”) security assurances. The blatant disregard of these commitments by Russia in 2014 and then again in 2022, and the cautious response by the other NWS, has inevitably raised questions about the value of such commitments.

The debate about Ukraine is further troubled by discussions about a different kind of security guarantee: that which results from the provision of weapons and systems, including in the long run, for the purposes of self-defence. This too is understandable: security is ensured as much by the constant access to arms as it is by diplomatic language, whatever its form. To wit, three key countries, Israel, Saudi Arabia and Taiwan, are de facto (but not de jure) allies of the United States because of the combination of strong language committing US assistance in case of aggression and sustained deliveries of defence equipment. 

While, on the basis of promises already made, Ukraine will eventually become a member of both NATO and the EU, thus benefitting from both organizations’ treaty-based security guarantees (Article 5 of the Washington Treaty and Article 42.7 of the Lisbon Treaty), membership will probably not come swiftly, and certainly not until the war can be said to have concluded.

So what to do in the meantime?  

After Vilnius: options before membership

Giving an Article 5-equivalent security guarantee to Ukraine immediately is politically impossible. No Western country will commit itself to defend occupied territories: that would mean deliberately going to war against Russia. Limiting the defense commitment to those regions which are under Kyiv’s control – a proposal sometimes made by Ukrainian officials before 2022 – is hardly feasible either: the battlefield is fast-moving and Russian air attacks continue far beyond the frontlines, including on distant cities.

In Vilnius, the allies simply stated that they “will be in a position to extend an invitation to Ukraine to join the Alliance when Allies agree and conditions are met” and that “the security of Ukraine is of great importance to Allies and the Alliance”. NATO and its member States have committed themselves to continue supporting Ukraine and a NATO-Ukraine Council was created, but no stronger language could be adopted. Separately, and responding to Ukraine’s disappointment at the ambiguity of the indefinite language in the NATO communique, the G7 declared that “the security of Ukraine is integral to the security of the Euro-Atlantic region” and announced that various bilateral and multilateral arrangements would be set up to ensure enduring assistance to Ukraine.

A durable ceasefire – should Ukraine and Russia ever agree to one – would leave open the possibility of a security guarantee even if Kyiv had not recovered its full territorial integrity. Absent accession to the Washington Treaty, such a guarantee could be given in different forms complementing each other. NATO could state that Ukraine’s security is “vital” or at least of “direct and material concern” to their own security (a language already used by the Alliance in 1991).

Separately, the three Western nuclear powers, the United States, the United Kingdom and France, which are also permanent members of the UN Security Council, could give guarantees akin to those given to Sweden and Finland once their candidacy to NATO was accepted, to avoid any attempt by Russia to change the status quo before these countries were formally covered by Article 5. 

Do precedents exist? Germany and Korea were divided after having been occupied by outside powers, leading to the creation, in each case, of two independent States. Also, the US guarantee to West Germany applied to a clearly identified territory (that of the then-Western Länder).  

A more interesting precedent is the language used in the US-Japan, US-Korea and US-Philippines treaties. They specify that US protection applies to territories “under the administration” or “jurisdiction” of protected countries. Thus, on paper, Kyiv could uphold its claims on occupied territories and be protected by Western guarantees on those it effectively governed.  

An even more useful parallel might be Cyprus: even though it is considered by the European Union as partly occupied since the Turkish intervention of 1974 and the creation of the Turkish Republic of Northern Cyprus, Nicosia does benefit, on paper, from the security guarantee of Article 42.7.

Finally, giving security guarantees to Ukraine should not leave other countries in limbo. Georgia, to which membership of NATO was promised in 2008, and Moldova, a neighbor of Ukraine, should also benefit from particular attention.

There will always be doubts about the credibility of any guarantees, in whatever form they take, but that does not mean they are worthless. As the record of NATO’s Article V testifies they can have a powerful deterrent effect. Those issuing the guarantees must maintain the capabilities to back them up. They also require regular reconfirmation. Questions have been raised for example as whether a future US President, whether Trump or someone with similar views, would feel bound by any promises made to Kyiv. This has led to suggestions that it might make sense for Congress to pass legislation comparable to that passed in 1979 with regard to Taiwan which requires the president to make available “such defence articles and defence services in such quantity as may be necessary to enable Taiwan to maintain a sufficient self-defence capability”. 

No country will wish to be wholly dependent on external guarantees for its security. This is why for Ukraine the important outcome of Vilnius was the commitment of extra resources and weapons from the countries present to sustain it in its war to liberate its territory from Russian occupation as much as the uncertain pathway to full membership of NATO.