The post-Cold War security architecture in Europe has been crumbling for the last decade and certainly has been shattered since Russia’s invasion of Ukraine in February 2022. Whilst much attention, analysis and discussion focus on the current war between Russia, Ukraine -and the indirect rivalry between NATO and Russia- there is also a need to think ahead of how future institutional arrangements can contribute to a degree of stability after the war will have ended. This GGI briefing paper argues that the current rivalry between NATO and Russia, most concretely expressed in the current Russo-Ukraine War, may need to be stabilized in part by a continental-wide, post-war conventional arms control (CAC) agreement. Whilst the duration and eventual outcome of the current war is of course far from predictable, such a future CAC agreement should improve diplomatic relations, reduce tensions, arms racing, and the likelihood of conflict – as such agreements have done throughout Europe in the past 100 years.
In November 2023, NATO member states suspended their participation in the 1990 Conventional Armed Forces in Europe (CFE) Treaty, following Russia’s 2007 suspension and in response to Russia’s final withdrawal in November 2023. Although the treaty had continued to function after Russia’s 2007 suspension between NATO members with and between non-NATO states such as Moldova, Belarus, and Caucus states, the treaty’s original intent was to limit forces between then-NATO and Warsaw Pact members and to reduce the likelihood of either launching a successful surprise attack. Attempts to adjust the agreement or come to a new Europe-wide agreement following the Warsaw Pact and Soviet Union’s dissolution have been minimally successful. While state parties revised the CFE Treaty in 1996 (the so-called ‘Flank Agreement’), all other attempts have failed. The 1999 Adapted (A/CFE) CFE Treaty was a comprehensive attempt at addressing the changed alliance structures and reduced threat, and although it was signed by all CFE Treaty state parties, it did not enter into force due to disputes over Russian military presence in Moldova and Georgia. Subsequently, other discussions and proposals even as late as December 2021 have gained minimal, if any, traction.
With a war raging in Ukraine amidst the first time in history that NATO and Russia are in conflict with one another, albeit indirectly, the need may come in the near future to stabilize the rivalry to prevent another, perhaps broader and more destructive conflict.
This briefing paper will provide an overview of adversarial CAC agreements’ purpose and goals; suggest why they are relevant today and in the near future; and summarize important issues relevant to future CAC agreements in Europe.
The overriding purpose of all adversarial CAC agreements is to stabilize the security relationship between adversaries. Stabilizing a security rivalry through arms control has many benefits including reducing the likelihood of conflict, improving diplomatic relations, reducing arms racing, and, as applicable, ending conflicts. Since World War One there have been approximately 30 CAC agreements in Europe, several of which are still in effect and approximately half of which were somewhat or fully successful. These agreements have assumed two, broad formats: limiting national capabilities of one or all parties to establish a fixed military balance or establishing an area of limited or no military capability, referred to as geographic demilitarization, to reduce competition and rivalry over a specific area such as an island or border area.
National limitation agreements such as the CFE Treaty mentioned above, the 1922 Washington Naval Treaty, or the 1936 Anglo-German Naval Agreement set quantitative limits on particular aspects on state parties’ military capabilities including capital ships, combat aircraft, and main battle tanks. These agreements stabilize military capabilities at a fixed ratio, reducing the need for states to acquire capabilities to keep up with one another. Other agreements such as the post-World War One and Two peace agreements set limitations on defeated states’ military capabilities – leaving the victors to retain or obtain whichever military capabilities they wish while the defeated states are constrained from seeking revenge and reversing their defeats. Because of wars’ expense, victors often seek to downsize their militaries so that the imposed limitations (albeit still with the defeated states consent and signature) sustain the victor’s military superiority even as they downsize.
Geographic demilitarization agreements may designate a geographic features such as the Åland (Finland) or Spitsbergen (Norway) islands as an area in which there are no military capabilities by any state party. Their goal is to remove a potential area of conflict between rivals when these geographic areas only pose a threat if a rival militarily possesses them but are otherwise not essential to a state’s defense. Other geographic demilitarization agreements usually are the result of a conflict and are intended to separate belligerent parties, and are often referred to as buffer zones. Physical separation, sometimes with an international force such as United Nations (UN) peacekeepers within the buffer zone, reduces the likelihood of the adversaries fighting one another and have been successfully implemented in Cyprus and the Balkans, for example (but notably failed in Ukraine).
CAC agreements almost always reflect the existing military situation and balance. It reduces the extent to which a militarily stronger party retains its superiority or it may retain parity between parties or blocs, but it never reverses relative military power. CAC agreements are always, by definition, an agreement between state parties; that is, state parties must consent to the agreement by signature (and depending on the state party’s system of government, requires legislative ratification). Even defeated states sign CAC agreements, usually as part of a broader peace agreement, to end being the subject of military attacks and to obtain, potentially, some concessions that they otherwise would not obtain were the conflict to continue and their defeat greater. These two aspects of CAC agreements are important because they emphasize that demands to substantially alter the military balance and relative military power are unlikely to be accepted, signed, and enter into force resulting in continued instability, rivalry, arms racing, and potential or continued conflict.
Most CAC agreements are characterized by specific definitions, well-defined geographic areas of application, and quantitative limits as applicable. While it is impossible to foresee every potential area of disagreement, subjectivity, or alterations in the geopolitical environment, CAC agreements usually strive to be as specific as possible to prevent state parties from cheating and obtaining an unintended military advantage.
One of the first questions to be addressed would be the issue of state parties – which states need to be signatories, which might be constructive inclusions but not necessary, and which might be wholly ancillary. At a minimum, all NATO states and Russia would need to be part of any agreement, with NATO representing a single bloc. Non-NATO EU members should likely be included because they have demonstrated non-neutrality as either direct supporters of Ukraine or indirect through EU policies. Moreover, collective defense articles in the Treaty on European Article 42(7) and the Treaty on the Functioning of the EU Article 222 for mutual defense and solidarity also suggest that, as EU members, their military capabilities should be counted. However, Central Asian states who were CFE Treaty signatories who may not need to be part of any future pan-European agreement.
There are several possible approaches to a future Europe-wide CAC agreement that, like the CFE Treaty, stretches from the Atlantic to the Urals (ATTU). The ATTU designation might refer to the state parties, to the area in which limitations apply, or both. First and foremost, any CAC agreement might incorporate, like the CFE Treaty, some version of continental zone limitations. With lower limits set along the common NATO-Warsaw Pact borders, ceilings increased the further one was from central Europe. This approach may ignore national borders (unlike the CFE Treaty which had a mixed zonal/national limits approach and A/CFE Treaty which had a national and territorial approach), viewing NATO and Russia as singular entities within which the zones would fall. While Russia previously objected to having limits within its borders, a zonal approach might likewise impose limits across NATO member national borders – offering Russia a sense of equal treatment.
A future agreement might return to the A/CFE approach of national and territorial limits, but this approach as well as others are complicated by the fact that NATO, especially its eastern flank, are composed of four states that border Russia (excluding Kaliningrad), with another two that border Belarus. Thus any agreement which imposes national and territorial limits on the Baltics in particular (who are geographically small and maintain small militaries) but do not impose similar limits on areas of Russia facing them would leave those three states vulnerable to attack and conquest. In sum, Russia would be able to mass forces within treaty limits against any one or even several NATO members, while national and territorial limits would prevent those states from being reinforced in response.
There are already geographic demilitarization agreements in place between NATO members and Russia – at least two of which predate NATO’s creation. The 1922 Åland Islands Convention and the 1936 Montreux Convention for the Turkish Straits are two agreements in which NATO members and Russia have agreements concerning limiting or prohibiting military capabilities. Other land areas including a buffer zone along the common NATO-Russian border, islands, or sea areas such as the Arctic or Baltic Sea might be subject to restrictions or limitations. Limitations might even apply to an entire state prohibiting, for example, either NATO or Russia from basing military forces in a “neutral” state.
Many CAC agreements limit specific military capabilities, or treaty limited equipment (TLE), such as armour, personnel, combat aircraft, attack helicopters, fortifications, naval vessels, and missiles – while leaving others untouched. Each agreement is in part a product of existing perceptions of threats and technology. While demilitarization agreements can be all-inclusive in restricting all military capabilities within a specific geographic area, balancing agreements require quantitative, qualitative, and definitional specificity.
The CFE and A/CFE Treaties set mutual limits on armored vehicles, battle tanks, combat aircraft, attack helicopters, artillery, and in a politically binding protocol personnel levels – laying out detailed definitions of these categories with lists of which equipment models were included within the limits. The treaty left untouched a number of capabilities such as most naval forces, heavy bombers, surface-to-air missiles (SAMs), radar systems, and command, control, communications, intelligence, reconnaissance, and surveillance (C3ISR) aircraft.
Any future CAC agreement would need to determine which TLE to include – a determination that would have to be agreed amongst all state parties. The five TLE in the CFE Treaty are a starting point, but additional systems including technologies that have emerged in quantity since 1999 might also be considered such as drones, long-range precision strike weapons, non-ballistic hypersonic missiles, autonomous weapon systems, medium-range ballistic missiles, and cyber weapons.
How states implement a CAC agreement is no less important than what is limited, and state parties may choose from establishing a system that is essentially bilateral, with no or weak agreement executors with inspections, monitoring, and disputes handled directly between state parties; or states may establish an agreement executor such as an international organization to which they delegate substantial responsibility. The Washington Naval Treaty, the Montreux Convention, and the CFE Treaty are examples of agreements with minimal delegation, while the Minsk Agreement and Cyprus buffer zones are examples of substantial delegation with the Organization for Cooperation and Security in Europe (OSCE) and UN being the primarily implementers of the two agreements, respectively.
Implementation approaches have traditionally been secondary to what is controlled and where and need not be a primary consideration of a CAC agreement’s framework but will, at some point, need to be negotiated and agreed.
Another major consideration, which may pose the greatest obstacle to an agreement, is determining the relative balance of forces. This discussion, for example was at the heart of disputes between the US and USSR during the MBFR negotiations. The Warsaw Pact’s overwhelming, quantitative conventional advantage over NATO made Moscow reluctant to agree to parity as such an approach would result in substantially more reductions than NATO. Their position changed, however, during CFE Treaty negotiations and as a result the Warsaw Pact accepted much larger reductions with a result of NATO-Warsaw Pact quantitative parity. The 1996 Balkans CAC agreement set a non-equal ratio between state parties but essentially retained the existing ratio of military equipment for the same five TLE categories as the CFE Treaty.
Thus, any continent-wide balancing agreement (in contrast to a demilitarization agreement) at some level needs to determine what the ratio of forces overall might be and, to an equal extent given geography, what ratio of forces along the common border areas might be (if this will be specified at all). Both sides might approach negotiations with claims to meriting a greater ratio; Russia because it might claim that it faces a strategic, long-term disadvantage economically, demographically, and as a measure of total defense expenditures while NATO might claim it deserves an advantage because of time-distance challenges of reinforcing its eastern flank and because it currently enjoys advantages for which it has no incentive to sacrifice.
This briefing has considered any Europe-wide CAC agreement as being separate from a settlement in Ukraine or linkages to nuclear agreements (which have traditionally been separated from CAC agreements). However, though it would likely complicate negotiations with the addition of more variables, a “package” agreement has the advantage of dealing with a greater number of NATO-Russia security disputes and may assist in ending the Russo-Ukraine War sooner.
Three important axioms about CAC which compel policymakers to consider the issue now are that: 1) CAC agreements often follow armed conflicts in Europe; 2) CAC is an important tool in stabilizing rivalries; 3) considering CAC goals and proposals in advance is more likely to result in an effective agreement and one that might be more quickly agreed upon. Moreover, the overlapping memberships and differing priorities amongst NATO and EU members will enormously complicate establishing common positions and proposals. CAC agreements are often characterized by specificity of limitations and geography, and NATO-EU members would be well served to harmonize positions earlier rather than later.
With these issues in mind, we recommend the following:
Photo credit: EPA-EFE/RUSSIAN DEFENCE MINISTRY PRESS SERVICE HANDOUT
