CIAO DATE: 02/02

The CFE Compliance Record a Decade After Treaty Signature

Kàroly Banai
Head of Section for Political Cooperation, NATO-WEU Dept., Hungarian Ministry of Foreign Affairs
and
Pál Dunay
Director, International Training Course, Geneva Centre for Security Policy

1999 - 2000

The Geneva Centre for Security Policy (GCSP)

Introduction

The 1990s turned out to be the decade of structural conventional arms control in Europe. Contrary to the decades of the post-World War II era, several agreements on the limitation, reduction or the outright ban of certain types of weapons were signed. Their scope ranges from the 1990 Treaty on Conventional Armed Forces in Europe (CFE) to its revised and adapted version of 1999 which marks the two ends of the process. Others include the 1996 agreement on sub-regional arms control under the Dayton peace accord and the anti-personnel land mine ban achieved in 1997. This paper deals exclusively with the compliance issues of the CFE Treaty. It aims to give an overview of the compliance record since 1990 and to look briefly forward to prospects of compliance in the future.

The achievements of the CFE Treaty are widely known. It was concluded by the members of NATO and the Warsaw Treaty and aimed to reach a balance of conventional forces in Europe between the two groups of states parties following a forty month long reduction phase. The five treaty-limited categories of armaments and equipment dealt with in the treaty are battle tanks, armoured combat vehicles, artillery pieces above one hundred millimetres calibre, combat aircraft and attack helicopters. In the case of the three land categories of armaments, regional limitations were introduced, whereas no regional limit applies to aircraft and helicopters. An agreement signed in 1992 supplemented the Treaty with limitations on personnel strength.

It represented a significant new element that the CFE Treaty included extensive associated measures as far as both information exchange and inspections are concerned. The huge amount of information extending to every piece of treaty-limited armaments and equipment of the armed forces identified with the military unit it belongs, serves as the basis of verification. The unprecedented level of transparency provides information that many states parties do not regard as particularly relevant under "fair weather" conditions. The verification measures, primarily on-site inspections, provide evidence on whether the information has been accurate. Even though the treaty refers to two types of verification measures, on-site inspections and national technical means, during the implementation a nearly exclusive reliance on the former was noticeable. As the verification is set against the submitted information, there is an intimate relationship between the two measures.

The CFE Treaty signed in 1990 is probably the last multilateral arms limitation agreement in Europe where, during the talks, adequate attention was paid to associated measures to guarantee compliance with the treaty, new ideas were explored and codified. To put it differently, it was the last one that was based on the existence of threat and mistrust. The advantage of this is that should relations among the parties deteriorate, for any reason, the stringent rules of the treaty would still be applicable. In case the relations are as good as they were most of the time during the first decade of its implementation, the treaty gives the possibility to carry out the implementation measures co-operatively.

It should be noted that there is neither compliance, nor non-compliance without the positive rule of the treaty. Consequently, the compliance record under the CFE regime can be measured only against the positive regulations of the treaty. It is necessary to emphasise two factors in this respect. There are two regulations one can not find in the CFE Treaty. First, there is no rule on circumvention. Consequently, any deviation from the rules of the treaty shall be regarded as a violation, i.e. as non-compliance. Second, there is no rule that would make certain types of violations "technical", that is, strategically insignificant ones. This would make it possible to separate the strategically insignificant violations from major ones. Again, in the absence of such a rule each activity or omission, which is not in strict concord with the treaty, is a violation. Consequently, the legal conclusion concerning compliance with the treaty may result in a far more rigid system than the one based on strategic and political assessment. Whereas under the former it can be concluded that there were a fairly high number of non-compliance cases, under the latter it is still possible to conclude that there is a nearly spotless compliance record.

It is a unique feature of the CFE Treaty that under a legally binding arrangement so many countries (22 upon signature, currently 30 states parties) carry out such a high number of on-site inspections. Since the entry into force of the treaty, approximately 3800 on-site inspections were carried out. We might discuss how many left some ambiguity concerning compliance. Instead of entering into details, it will suffice to state that, according to NATO estimates, no less than 94 per cent of all inspections left no doubt that the inspected party complied with the treaty. If this assessment is correct, one can conclude that the compliance record is fairly good. During the first decade of treaty implementation, due among others to the changed structure of security relations, the practice of co-operative verification was established. States parties have very often carried out joint inspections, thus sharing the costs and benefits of this activity.

One can certainly attribute some importance to the fact that no large-scale violation of the treaty has occurred in the decade that has passed since the signature of the CFE Treaty. These years, except maybe for 1991, have not been dominated by compliance issues. The focal point of attention has been the constant adaptation of the treaty to fast changing circumstances. Suffice it to mention a few major steps in the process, such as addressing the withdrawal of Soviet armaments and equipment behind the Urals, the reassignment of pieces of treaty-limited armaments to naval forces not limited by the treaty, the status of the so-called flank area after the dissolution of the Soviet Union or coping with the first eastern enlargement of the Atlantic Alliance. In light of such fundamental re-arrangements of the strategic situation, fairly little political attention has remained for the compliance record. This has certainly been favourable since the states parties could carry out their activity professionally without high level political interference. Thus, even some highly controversial matters could be resolved. This was further facilitated by the fact that the treaty established a number of institutions to address non-compliance and other matters of implementation. The Joint Consultative Group, a relatively low profile body, has held sessions in Vienna as necessary. Extraordinary conferences could be convened on a number of occasions to address issues that required some innovative approach or that would have made formal adaptation of the treaty necessary. Last, but not least every five years the review conference of the states parties can be convened in order to address compliance matters and revise the treaty as deemed necessary. The first review conference in 1996 adapted the flank rule to the changed circumstances and in spite of the signing of the adapted treaty in November 1999 some countries will continue with the adaptation process at the next review conference in 2001.

Considerations To Analyse Compliance

In order to come to any sound conclusion concerning the compliance record under the CFE Treaty, it is necessary to establish categories that make this possible. The analytical framework, according to my impression, should be established on four grounds: 1. The cause of violation. In this respect a difference can be made between non-compliance due to the a) Change of the strategic importance of a part of the area of application; b) Need to satisfy the military needs of a pending (burning) conflict; c) Inability to carry out the reductions set forth under the treaty. 2. The time of the activity that resulted in non-compliance. In this respect it is necessary to differentiate between three phases: a) Events that occurred before the signature of the CFE Treaty; b) Events that occurred between the signature and entry into force of the CFE Treaty; c) Events that occurred after the entry into force of the treaty. 3. The state party that has committed the violation. 4. The attitude of the state party not complying with the treaty: a) Recognition of the treaty violation; b) Denial or non-recognition of the violation. Each of the four approaches may prove to be a useful tool to draw conclusions concerning the compliance record under the treaty.

The analysis of the compliance record will be carried out with a focus on the cause of violation. It will also be assessed how the different types of non-compliance cases can be categorised according to the other criteria mentioned above.

The First and Most Complex Test of Compliance With the Treaty

Even though, as will be demonstrated later, several cases of non-compliance with the CFE Treaty turned out to be complex, the most complicated one, according to the criteria established above, was the very first case the signatories of the Treaty faced. The day before the CFE signature in Paris, the states parties exchanged information, as required upon the formal conclusion of the talks in Vienna. The Soviet Union notified a surprisingly small number of treaty-limited armaments and equipment upon signature, had withdrawn thousands of pieces of armaments, had reassigned three divisions to naval infantry and last, but not least, did not notify the armaments and equipment of four formations and units. As the Dutch ambassador to the talks put it later "... on the eve of those euphoric days of the Paris Summit ... the countries which had participated in the Treaty negotiations were in for a cold shower." The first papers published after treaty signature seemed to be more concerned about the armaments withdrawn behind the Urals than the two other steps taken by the Soviet Union. According to my impression, signatories should have been less concerned about the arms deployed in Asia than about the increased capability of Soviet naval infantry in Europe. The equipment withdrawn behind the Urals was several thousands of kilometres away from any potential area of confrontation in Europe. Thus the violation had nothing to do with the objectives of the CFE talks, namely "the elimination ... of the capability for launching a surprise attack and for initiating large-scale offensive action" - as set out in the mandate of the CFE talks.

Concerns were also legitimate with respect to the three divisions reassigned to the naval infantry. The primary reason for concern should not be the amount of reassigned armaments, although it was not negligible either. According to reports, the shifting of three motorised rifle divisions to the naval infantry involved the reassignment of 807 battle tanks, 896 armoured combat vehicles and 838 artillery pieces. What puzzled most signatories was that the "three divisions in question were, until recently, army divisions, and from the explanations given by the Soviet Ministry of Defence, it appears that they continue to be land force divisions, without any amphibious capability or role. They are therefore land force formations or units, whose equipment should count under the numerical ceilings, and with the units themselves constituting objects of verification." This situation carried the danger that if parties had accepted such an approach it could have served as a precedent. "Naval forces", not subject to the limitations of the treaty, could have appeared on the Western border of the Soviet Union, reinforcing armaments subject to the limitations of the treaty. Theoretically one could also imagine, in connection with the thousands of pieces of armaments withdrawn behind the Urals, that they had been reintroduced in the area of application and assigned to naval forces.

The Soviet Union put forward several arguments to legitimise its position. It referred to the rules of the treaty under the mandate of the CFE negotiations, according to which "naval forces ... will not be addressed", and to the argument that the area of application extended from the Atlantic to the Urals. Furthermore, it maintained that the reassignment, or as it was sometimes called the re-subordination, took place well before treaty signature in 1989 following a decision of the military, having nothing to do with the CFE talks. The purpose of the re-subordination was to ensure more reliable protection for Soviet sea lanes in view of the perceived considerable advantage held in naval systems by the West. At this early stage after treaty signature, other signatories took a firm stand. The U.S., for example, held the view that "the CFE treaty should not be sent to the Senate for ratification until the 'naval defence' dispute was settled".

In order to overcome the stalemate, and after several compromise proposals, the U.S. Secretary of State and the Soviet Foreign Minister reached a compromise on 1 June 1991 in Lisbon. This was followed by an announcement of the Soviet Union at the JCG. According to this, the Soviet Union committed itself to destroying at least 6,000 battle tanks, 1,500 armoured combat vehicles and 7,000 pieces of artillery from the armaments behind the Urals in addition to its commitment to reduction in the European part of the country. Another part of the armaments withdrawn from Europe behind the Urals was promised to be placed in storage or used for replacement and repair in that part of the country. Those pieces of armament were promised not to be used to create a strategic reserve or operational groupings. No access to the armaments behind the Urals was granted. Thus, countries not having national technical means at their disposal were not in the position to assess how the Soviet Union, and later Russia, maintained the withdrawn armaments. Such a situation materialized years later, in 1995, when in a trial inspection under the Open Skies Treaty, German experts learned that thousands of withdrawn pieces of armament were exposed to the Siberian climate. It is clear that their military value has been reduced to zero during the years. The Soviet Union also committed herself not to continue the reassignment of armament to naval units and rocket forces, whereas the West was ready to accept the pre-signature reassignment as a fact.

The complexity of this case stemmed not only from the cumulative references to various grounds but also from the timing of the Soviet activity. It was not in concord with the Treaty and had started before treaty signature. It was clear that the strategic and political assessment of Soviet behaviour and the legal evaluation of the situation could not be the same. Whereas it was clear that tolerance towards the Soviet activity could undermine the strategic rationale of the treaty, stricto sensu such activity could not be regarded as a formal treaty violation. This could be one of the reasons why the problem was solved by compromise. The other, far more important reason, was that since the signature of the treaty the Soviet Union and its armed forces have remained the single most important motive to keep the CFE in place. In such situations it is important to engage the party whose military capability is the decisive factor in the arms control arrangement. Otherwise stated, it is important that the other parties assess carefully what needs to be done in order to obtain compliance from the state in question, but without alienating it to the extent it would consider withdrawing from the treaty. It is also interesting to note that this was one of the very few cases when a party, not entirely without reason, denied that it had violated the treaty.

Change of the Strategic Importance of Part of the Area of Application

As mentioned above, the first decade of existence of the CFE Treaty was dominated by adaptation to changing circumstances. Among the causes of this, the dissolution of the Soviet Union played a prominent role. Areas that had no strategic importance unexpectedly became important for two successor states, Russia and Ukraine, which did not deny their reluctance to respect the limitations of the treaty, and sought to revise them. Contrary to the above mentioned re-subordination and withdrawal of treaty-limited armament, instances in this case, the overwhelming majority of states parties showed understanding to the claims of the two countries, even if they did not agree upon the solution recommended by Russia and Ukraine.

While the overall compliance with the Treaty in general terms, and with the reductions in specific terms, has been overwhelmingly good, the same cannot be said about the flank area. The treaty established a "combined" flank zone that consisted of a northern and a southern part. This gave some flexibility to Russia (but not to Ukraine) to move treaty-limited armaments from the stable North to the turbulent South as necessary. Following the dissolution of the Soviet Union, it was clear that Russia would exceed the flank limits set by the Treaty, unless it could carry out significant relocation of treaty-limited armaments. Russia opposed it on a number of grounds. 1. Sometimes it was based on the recognition that the military situation was too unstable to respect the treaty limits in the area, for instance when President Yeltsin wanted to exempt TLE in Chechnya. 2. It was also argued that the disadvantageous change of the strategic situation could only be compensated by further reductions by the smaller newly independent Transcaucasian republics. 3. Last, but not least, the high costs of putting the reductions into practice through the relocation of troops was put forward as an argument.

It seems that despite the lasting violation of the flank limits set forth by the treaty, the attitude of other states was, with the exception of directly affected parties from time to time, co-operative and relaxed. They showed understanding towards the problem and comprehended that the way out of the situation was not exclusively through coercion, but through a complex set of measures. It was noted that the non-compliance with the treaty limits on the flank (primarily in the south) had not reached strategic proportions, but remained limited. When the reduction process had to be completed in November 1995 the Joint Consultative Group went so far as to state that Russia was in "technical non-compliance".

The steady non-compliance with treaty rules on the flank offers the following conclusions. The violation was due to the changing strategic value of certain parts of the area of application. This was aggravated by the recurrent domestic instability of the Southern flank that, by the same token, gave some legitimacy to the violation. The continuous adaptation of the flank rule compromised the compliance effort. Both Russia and Ukraine were ready to admit their non-compliance. This also served as an attempt to generate the understanding that their activity or, for that matter, omission was not directed against any other state party, thus aimed to create some confidence of other parties in the non-compliance process.

It is a fact that most recently, in the adapted CFE Treaty of November 1999, the parties eliminated the flank rule and the special flank limitations have been inserted in the Protocol on Territorial Ceilings.

Need to Satisfy the Military Needs of Pending or Burning Conflicts

The CFE Treaty is a document of peace. By taking a closer look, it seems clear that it does not deal with violent military conflicts, or with war for that matter. Several rules of the treaty can be mentioned in this respect. The simplest and most convincing example may be the so-called reduction protocol. The detailed and highly technical document fully ignored the possibility of war in the area of application. It listed, for example, several methods of reducing a piece of armament or equipment that did not include destruction by an adversary or war. In spite of this, destruction by war occurred in some of the successor states of the former Soviet Union, namely, in Azerbaijan, Georgia, Moldova and Russia. War challenges the foundations of the CFE Treaty. Fighting a large-scale conventional war does not usually result only in the violation of international law generally, but also of the CFE Treaty specifically. It requires force concentration that may result in exceeding the limitations imposed by the treaty in case the territory of a state party is not confined to one zone of the area of application. It results in certain constraints on transparency and makes on-site inspections in the war zone at least temporarily impossible. In the case of inter-state conflict, or in the more frequent case of an intra-state conflict with the involvement of another state party, the situation is complicated further. As wars have taken place on the flank, whose strategic importance changed significantly it is also related to the above matter and appears cumulated with it.

The wars in Chechnya had the very same effect on the CFE Treaty. 1. In both cases Russia exceeded the flank limit in the three treaty-limited land armament categories. 2. During the hostilities no reliable information exchange was possible partly because of the "fluidity" of the situation, partly because of Russian reluctance to provide for it. 3. For security and personal safety reasons, during that period it was impossible to carry out on-site inspections in the area where military action was taking place. When the war is over inspections can restart. Access to the area generally and on-site inspection specifically, is denied more often in post-war situations than otherwise. The information provided unilaterally by the warring party cannot be checked through on-site inspection. Thus, it remains difficult for a while to get reliable information about compliance. A blank spot in the implementation of the treaty is created.

In case of the first Chechnya war, even if some parties complained about Russian intention to exceed limitations in three land categories on the southern flank, the Russian demand was settled by compromise thanks to the approaching review conference where the flank rule was due to be revised. Thus, only those countries that felt directly affected threatened to introduce counter-measures. During the second war in Chechnya, Russia provided at the Joint Consultative Group with information on 19 October 1999. It announced (on 10 October) to have 1493 battle tanks, 3534 armoured combat vehicles and 1985 artillery pieces on its flank territory. It had exceeded the flank sub-limit, as agreed by the negotiating group on 30 March 1999, by 193 battle tanks, 1394 armoured combat vehicles and 305 artillery pieces. After the achievement of the most important military objective, the occupation of Grozny, Russia indicated its willingness to station a division strength unit north, and a regiment strength one south of the Terek river in Chechnya. Assuming that Russia will not reduce the heavy weaponry of its formations in Chechnya one has to take into account the excess of the flank limit in the long run. On July 1 2000, the excess armament in Chechnya was reduced to 150 battle tanks, 885 ACVs and 317 artillery pieces. This is still a significant violation of the flank rule, though it is less in two categories than three months earlier. The motivation of Russia seems clear. On the one hand, Russia wants to demonstrate that it intends to comply with the treaty. On the other, however, it does not want to jeopardise its security interest. It is self-evident that no country will give priority to compliance with arms control when it feels its perceived security interest is at stake. Or, as Russian officials have stated, "...they will not comply with treaty limits, until Moscow completes its self-described anti-terrorist operations in Chechnya". Inspections in the region, with reference to safety and security problems, will continue to be rejected for some time by Russia. This, in turn, increases the concerns of other states parties and results in a delay of the implementation of the adapted treaty provisions.

Two other conflict zones, Georgia and Moldova, have some elements in common. They are two sovereign successor states of the former Soviet Union where, due to some internal conflicts, a certain number of foreign (Russian) troops are stationed together with their treaty-limited armaments. It is clear that such stationing cannot take place without the consent of the party on whose territory it takes place. This stems from the principle of state sovereignty, among others. The revision of the CFE flank rule at the first review conference of the Treaty in 1996 resulted in a situation that Russia could refer to as legitimising its presence on the territory of other former Soviet republics on the flank. It formulated the revised Russian reduction obligation with reference to "the area ... as understood by the Union of Soviet Socialist Republics at the time the Treaty was signed". Russia, following the adoption of this text, has given the same reply whenever the issue emerged in the JCG. It maintained that the issue did not pertain to the subject matter of the CFE Treaty and should be tackled bilaterally. Thus Russia was ready to recognise the existence of the problem, it did not deny the presence of Russian armaments on the territory of the two countries, but argued that the forum where it should be discussed. The above approach contained an element of truth, as the problem was related to pending internal conflicts in the case of both Georgia and Moldova, which Russia provided with military assistance. The matter, however, had a lot to do with the application of the CFE Treaty, however, as Russia in both cases stationed hundreds of pieces of treaty-limited armaments without the gradually diminishing consent of the territorial states. Because of the regular denial of inspection rights, it was extremely difficult to collect reliable information about the situation in the relevant parts of Georgia and Moldova. The formal solution was partly achieved at the OSCE summit in Istanbul in November 1999, at least in one of the two cases. Russia took the commitment to withdraw its troops from Georgia. One has to bear in mind that not long after the "formal regulation" of the matter, Russia criticised Tbilisi for not being able to control the common border between Chechnya and Georgia. In light of this, it remains open to question whether Russia will live up to its commitment. Not to mention that the conflict in Abkhazia, if not in Ossetia, may pose a threat to the unity of Georgia. As of the summer of 2000, one can state that after some tough bargaining, the Russian withdrawal from Georgia has begun. Russia makes apparent attempts to delay the elimination of the presence of the strategically important air force in Georgia and gives priority to the implementation of the withdrawal of ground forces. It remains to be seen whether for one reason or another (e.g. eventual blackmailing of Georgia, lack of political settlement in Abkhazia) the multi-year withdrawal process will be put to a halt or compliance with the bilateral Georgian-Russian accord will prevail.

In the case of Moldova, despite the regular claims of Chisinau, Russia offered even less. It kept the withdrawal of Russian troops conditional on a political solution concerning the status of Transdniestria. In this case it is even more doubtful whether Russia will be ready for the withdrawal of its forces. Russia has often put forward the argument that the withdrawal of its troops from the territory would undermine stability and referred to the logistical problems of transporting huge amounts of ammunition through Ukraine. The U.S., in reaction to this, suggested at the beginning of 2000, for the first time, that it could make 30 million USD available for the purpose of withdrawal if the local leadership in Tiraspol (Transdniestria) behaves co-operatively. Moldova, on a number of occasions, referred to the possibility of not ratifying the adapted CFE Treaty, in case the Russian withdrawal from its territory does not take place.

The case of Russian stationing in Armenia was different from the Georgian and Moldovan cases in two respects. 1. Whereas the latter two states have successfully multilateralised their problems with Russia, through the CFE adaptation talks, Armenia did not. Multilateralisation meant both the establishment of the so-called GUAM group (Georgia, Ukraine, Azerbaijan, Moldova), made of countries that have had somewhat similar problems with Russia, and the pursuing of the discussion in the framework of CFE fora. Armenia kept the issue of Russian treaty-limited armaments bilateral. 2. In the case of Georgia and Moldova, Russia got involved in the domestic conflicts of the two countries, in the case of Armenia, in an inter-state one involving Azerbaijan. 3. In the case of Georgia and Moldova the so-called unaccounted treaty-limited equipment (UTLE) problem appeared on the margins, while with respect to Armenia it played a central role.

Russia has favoured Armenia's position during the conflict concerning Nagorno-Karabakh. It has co-operated closely with Yerevan, deploying troops in the region, including battle tanks, ACVs and heavy artillery. Azerbaijan, in order not to face inferiority vis-à-vis the Armenian forces (assisted by Russia), violated some limitations in turn. The secret Russian armaments supply to the Armenian troops stationed on the occupied territory of Azerbaijan took place between 1994 and 1996 and was later acknowledged by the then Russian Minister of Defence, Rodionov. It included 84 T-72 battle tanks, 50 BMP-2 armoured infantry fighting vehicles and 72 artillery pieces of the D-1, D-20 and D-30 types. It is clear that the violation of the CFE Treaty was a symptom and not the cause of the problem. Without a lasting political settlement for the Nagorno-Karabakh conflict, lasting respect for CFE rules was inconceivable.

Inability to Carry Out Reductions Set Forth By the Treaty

Since the beginning of reductions under the CFE Treaty, some states parties have regularly referred to their inability to fulfil their obligations. In some cases the insufficient technical capacity, in others the shortage of financial means were factors in the problem. It was not unprecedented that a state party combined these problems with some political protests and/or issues of financial support. NATO members were in a better position than former Warsaw Treaty countries for three reasons: 1. NATO remained a cohesive alliance after the conclusion of the CFE Treaty and excess armament could be cascaded to member states that had less developed military technology. The same did not happen on the eastern side, where shrinking armament markets made the large-scale selling of such armaments largely (or nearly) impossible. 2. The East had to reduce much larger number of armaments. The only western state party to the CFE that had to carry out substantial reduction was Germany, due to the inherited arsenal of the East German Nationale Volksarmee. 3. The West did not face the severe economic decline the East did at about the time most reductions were due.

Remarks were made by a number of former Warsaw Treaty countries about their inability to carry out reductions in time. Belarus, Kazakhstan, Russia and Ukraine were among them at the end of the reduction phase. Minsk and Kiev were in the most difficult situation as they had inherited a disproportionate part of Soviet heavy weaponry. Belarus requested 12-15 million USD to complete its reduction obligation. Later it sold some pieces of armaments to Bulgaria and Hungary. Whenever the issue emerged, other states parties had to find the necessary balance between perseverance and flexibility. So they were ready to extend the reduction period, for instance in 1997, when Belarus asked for a further extension of the period to reduce 185 armoured infantry fighting vehicles, while at the same time urging the party to continue with the fulfillment of its obligations. It is clear that, in spite of the delay, the reductions originally set forth by the CFE Treaty will be put into practice fully. As those states that have been lagging behind with the implementation are not planning to carry out further reductions under the adapted treaty, its entry into force will not add to the implementation problems. Those states that have taken unilateral commitments to carry out further reductions, in case the adapted treaty comes into force, have, in most cases, already started to turn their promises into practice. They have done it in anticipation of the new treaty, or to avoid burdening their defence budgets with the maintenance costs of armaments that should in any case be cut down in the near future.

Conclusions

The implementation of the CFE has been largely successful during the first decade of the operation of the Treaty. The overwhelming majority of inspections have not revealed non-compliance by any state party. There was no major discrepancy between the exchanged information and that acquired through on-site inspections. No state violated the Treaty in order to achieve decisive strategic advantage in its framework.

The only violations, with the exception of some ambiguity concerning the interpretation of the counting rules of the treaty after signature, were committed by successor states of the former Soviet Union without exception. This was due to several reasons that may illustrate the complexity of the issues in the post-Soviet space. The legacy of the Soviet Union includes a multitude of problems that sometimes have appeared cumulatively as grounds for non-compliance. Some of these are: 1. The comparative autonomy of military decision-making structures (the absence of clear subordination to political authority) that was the main reason the Soviet Union's treaty-limited armaments re-subordination and their withdrawal behind the Urals before treaty signature. It was a major contributing factor that the Soviet/Russian military were in favour of armed forces that loomed larger than their real strength. This was the case in many other countries where the political leadership was, for a shorter or longer period of time, mislead by the military leadership concerning the performance of the armed forces. 2. The inability to achieve domestic stability and the perception that large-scale military manoeuvres, like the ones conducted in Chechnya, would provide for the exclusive solution against aspirations to secede. 3. A related problem is how Russia perceives its immediate neighbourhood. It intends to be surrounded by like-minded regimes in the former Soviet space and claims the right to be a party to conflicts on their territory. The heavy reliance on military means, in many of those cases, interferes with Russian willingness to comply with the conventional arms control commitments. 4. It is also part of the military legacy that the carrying out of costly large-scale conventional arms reductions "ate up" the peace dividend due to the reduction of the size of the armed forces. In this case the three largest European successor states of the former Soviet Union were most heavily affected, particularly those western-most republics that inherited a disproportionate part of Soviet conventional armaments.

It is interesting to see that each violation was rooted in the Soviet past. Each solution, taken in the framework of the treaty, was without exception a compromise. Sometimes this meant that parties consented to the expressed or tacit revision of treaty regulations. In other cases at least a temporary, though lasting exception to treaty rules was accepted. If one agrees that most violations had their roots in the Soviet past and were not strategically threatening, such an attitude can be regarded as suitable. There are grounds to assume that three of the four above-mentioned reasons for non-compliance will vanish. Sooner or later the post-Soviet states will live up to the treaty requirements. Whether Russia will retreat from conflict zones in its neighbourhood or not is more difficult to anticipate. In such cases the violation of the CFE should be regarded as a corollary, requiring the elimination of the source of conflict rather than addressing the violation on the surface.

Russia's engagement in the conventional arms control process is valuable, particularly for those small and medium size European states that would be mostly affected by a Russian break-out of the system. It is essential to continue to carry out intrusive inspections in order to make any attempt to opt out of the system impossible or at least costly. The most important question, however, is to make it clear that if tolerance prevails with respect to non-compliance stemming from the legacy of the past, the same will not apply to violations due to the reinterpretation of a country's role, in the sense of neo-imperialist aspirations.

The territorial concentration of non-compliant countries is an indication of the high level of military stability achieved in Europe, outside Russia and its neighbourhood. Compliance with the CFE also illustrates that European states share the view that it is not the number of conventional weapons that will guarantee their security. Whether this is a demonstration of the irrelevance of conventional force limitations would be a conclusion prematurely drawn from lasting "fair weather" conditions. It must not be forgotten that the CFE Treaty does not extend to the former neutral and non-aligned countries, nor to the successor states of the former Yugoslavia. In light of the well-established rules and the extensive practice of its application, the CFE may be a tempting point of reference for Europe's other violence-prone crisis zones.