Carry that weight: assessing continuity and change in NATO’s burden-sharing disputes

ABSTRACT This article proposes Stephen Jay Gould’s concepts of time’s arrow and time’s cycle as a conceptual tool to analyse NATO’s burden-sharing disputes. It argues that the controversies on burden-sharing in NATO can be assessed in terms of their cyclic or arrow kind nature, rendering some disputes more likely to recur than others and providing different kinds of starting points for their forecasting. The study identifies four cyclic categories in which burden-sharing has transformed into a political debate among NATO members during the post-Cold War era: geopolitical change related to Russia; periods of US foreign political retrenchment or renewal; the passivity or activism of European NATO members; and during NATO or allied out-of-area operations. Moreover, the study suggests an arrow kind of direction in burden-sharing disputes, indicating an expansion of disputes to cover comprehensive security, resilience, security co-operation and diplomacy, and to engulf also NATO partner countries.


Introduction
Once again, we seem to live in times of NATO in crisis. While NATO's external security environment has during the recent years seen a deteriorating level of stability due to increased tension with Russia and the mounting unrest in the Middle East, the organisation is at the same time met with a painstaking transformation from a crisis management organisation, back to one focusing on collective defence. Moreover, tensions within the transatlantic community due to American global military over-stretch, and its growing frustration with the level of European contributions, have gradually intensified during the 2000s and turned into nearly an open quarrel since the start of the Trump Administration. 1 Question of intra-Alliance burden-sharing are in many respects at the core of these tensions. 2 However, the issue is nearly as old as the Alliance itself. While burdensharing has received substantial attention during and since the 2016 U.S. Presidential elections, it seems that disputes related to it emerge and re-emerge to the transatlantic agenda on a regular, if not predictable, basis. 3 In particular, burden-sharing tends to gain additional prominence during transitional phases of the Alliance. Events like organisational reforms, enlargement, major changes in dominant military strategic thinking (for example shifting emphasis on conventional vs. nuclear weapons in defence policy), or new out-of-area operations, have often led to intra-Alliance debate on how the new costs or additional responsibilities should be divided.
Burden-sharing disputes are thus simultaneously a recurring issue and something that is difficult to manage. Based on historical experience, the disputes may not end up in NATO's undoing or unravellingdespite occasional claims about "unfair" burdensharing, no ally has thus far withdrawn from the clubbut they nevertheless are apt to weaken the Alliance's internal cohesion and thereby its ability to meet contemporary and future challenges. In a longer term, such disputes may, along with other factors, contribute to the overall gradual withering of the Alliance.
The purpose of this paper is to contribute to the discussion on NATO burden-sharing by seeking to construct a conceptual framework on the dynamics of intra-Alliance controversies related to it. Using academic literature on NATO's post-Cold War era burdensharing disputes as empirical source material, the article will argue that the past controversies related to burden-sharing can be assessed in terms of their cyclic and arrow kind nature. In this respect, the article builds on Hartley and Sandler's and Driver's quests to identify trends or cycles of the phenomenon on the one hand and Hallams and Scheer's observation of NATO's burden-sharing debates as a repetitive phenomenon on the other. 4 Moreover, this article's ambition is to study the nature of burden-sharing disputes within NATO in order to facilitate their forecasting. The goal is not to produce an accurate prediction of a particular future or scenario; instead, the article's approach is more akin to a typological theory, which helps to identify trends and indicators related to burden-sharing disputes within NATO. Following George and Bennett, a typological theory can be understood as a theory that specifies independent variables, delineates them into the categories for which one can measure the cases and their outcomes. 5 A typology provides not only hypotheses on how these variables operate individually, but also contingent generalisations on how and under what conditions they behave in specified conjunctions or configurations to produce effects on specified dependent variables. In contrast to a general explanatory theory of a given phenomenon, typological theory seeks to provide a rich and differentiated depiction of the phenomenon and can generate discriminating and contingent explanations and policy recommendations. 6 Thus, this article does not seek to give final answers, but rather fosters a differentiated, policy-relevant way of looking into the issue at hand. Such an analysis of burden-sharing controversies within NATO may at its best be actionable knowledge: it helps to understand better the dynamics and vulnerabilities of NATO's internal coherence and thereby enable better indicators or mechanisms both within member states and NATO structures in order to manage or mitigate future disputes, possibly preventing some of them taking place altogether.
From a methodological point of view, the approach proposed in this article seeks to answer the calls for nuanced understanding of burden-sharing and the need for mixed methods. 7 It will be argued that in addition to diverse distribution models prevalent in academic studies of NATO, 8 the perspective of intra-NATO disputes sheds additional light to the question of burden-sharing. Metatheoretically, the article's perspective emphasises burden-sharing's nature as belonging to the domain of inter-state bargaining: even if decision-makers or academics were able to construct an "ideal" model of burdensharing, NATO members would not stop seeing the issues through favourable lenses for themselves. The emphasis on these two dimensions of NATO burden-sharing disputestheir cyclic and social nature, reflects the underlying understanding of this article that burden-sharing is a complex, multi-dimensional, and evolving issue. In fact, in the light of post-Cold War experience, the phenomenon as a whole may escape our attempts analytically to capture it, leading to the need for more limited empirical studies such as the current study.
This article is divided into three sections. Section on materials and methods provides an enquiry into the metatheoretical nature of the issues at hand, highlighting the sociallydefined political space in which the topic is manifested. It also discusses the logic of repetition and continuity in NATO's burden-sharing disputes. Moreover, the section reviews empirical research materialthe post-Cold War era academic literature discussing burden-sharing disputes within the Alliance. After that, a section devoted to results sorts out the conditions leading to these disputes based on past experience, concluding with four relatively repeating and cyclic categories related to Russia, the USA, European allies and NATO out-of-area operations, plus an assessment of the direction or the "arrow" of burden-sharing disputes. Finally, the section conclusions focuses on the promises and limitations of the proposed approach in relation to existing scholarly literature, as well as on possible avenues for future research.

Metatheoretical considerations of burden-sharing
What characterises this article's reading of NATO's burden-sharing disputes is the essential role of political considerations. Research has consistently shown that NATO members do contribute differently to collective defence and that these differences in spending tend to be long lasting. 9 However, public disputes on how to share the burden within NATO are not equally stable or constant. Instead, they owe their existence to a conscious decision by policy-makers in member countries. Whatever the material facts on the field may be, the imbalances on NATO agenda are governed by political logic. In other words, national decision-makers can always choose whether or not to turn public attention to the disparities or controversies amongst member countries. These considerations are affected by a set of changing domestic, economic, geopolitical, or technological developments, and evolving understandings of national interest in NATO members.
Accordingly, during the course of NATO's history, members have tended to emphasise those qualities of security in which they have contributed the most and to discount the value of the efforts undertaken by others. 10 This way, the USA for example has typically turned attention to the military dimension of transatlantic security, whereas many European NATO members seem to prefer to talk about wider issues of security, such as diplomacy, their role in peace operations, or their economic contribution to global and regional stability. These policy emphases can be seen in the light of efforts to influence the political space of NATO burden-sharing and to frame the debate in positive light for oneself.
Following these thoughts, this article's metatheoretical approach involves a fusion of realist and institutional accounts of international relations. The underlying assumption is that whilst alliances and interstate co-operation are sometimes difficult to achieve, and always difficult to sustain, because of relative-gains considerations and concern about cheating. 11 States nevertheless construct not just short-lived alliances, but more permanent institutions in order to overcome security challenges they cannot manage alone. 12 Institutions are generally costly to create but they are relatively cheap to maintain, which is why it is in a member states' interests to maintain an alliance, even after the disappearance of the original threat. 13 Institutions may mitigate disputes amongst participating states, but may also develop a degree of agency. In NATO's case, the Alliance has an integrated military command, a secretary general, and an international staff comprised of military and civilian personnel, even though NATO member states retain a high degree of control over the Alliance organisation when compared to for instance the European Union. As an outcome of the interplay of neoclassical realist and institutionalist logics, NATO burden-sharing disputes should be seen as a continuous struggle within a commonly agreed framework or, in other words, a regulated struggle, the agenda of which is evolving continuously.

Time's cycle and time's arrow
In its effort to model NATO's burden-sharing disputes, this article assumes its starting point from paleontologist Stephen Jay Gould, who in his 1987 book Time's Arrow, Time's Cycle discusses two schools of thought about geology in terms of their basic orienting metaphors. According to Gould, one school of geology sees the large-scale history of the Earth in terms of cycles in which there is change from one phase to another but the phases themselves recur through regular cycles; the other sees geology as revealing constant unidirectional change, like an arrow in motion. 14 The notions of arrow and cycle are a powerful pair of metaphors. Time's arrow captures the uniqueness and distinctive character of sequential events. Through the lens of the arrow metaphor, history is seen as an evolutionary and irreversible process which cannot be turned backusing the household analogy, if one mixes coffee with milk, there is no turning back and "unmixing" the two substances. Then again, the metaphor of time's cycle provides these events with another kind of meaning by evoking lawfulness and predictability 15returning to the use of substances, for many people, the use of coffee in everyday life tends to follow regular repeating patterns. Each perspective can have an element of truth, as Gould argues is the case for Earth's history, and we should be suspicious of any unqualified answer. But the question of the extent to which and the ways in which international history resembles a cycle or an arrow can be useful one also in terms of forecasting the future. 16 Indeed, does history have a direction of certain kindor is history more about repetition or enduring relevance of certain recurring patterns? This debate has been ongoing in the domains of history and social science for centuries by scholars such as Spengler, Pareto, Toynbee, and Sorokin with more contemporary interpretations provided by Schlesinger Sr. and Jr. in their cyclical theory of U.S. History. The question has of course also animated political scientists: a notable example is Paul Pierson's study on path dependency and notion of politics of time. 17 In international relations study, the main emphasis has been on power transition theories developed after World War II by authors such as Organski or Gilpin, attempting to explain how international orders tend to collapse by resorting to war. In turn, Modelski's theory of long cycles puts forward an evolutionary perspective of the global system, which is cyclically marked by systemic transitions. 18 However, the power cycle theory, introduced by Doran may be the closest equivalent to this article's approach. According to Doran, states have a pattern of experiencing a cyclic rise and fall from power. Doran's power cycle theory displays the cyclic nature of nations' roles in the international system and seeks to explain the effects of this cycle, especially in terms of war causation. To do this, Doran suggests the measurement of national material capabilities across a number of indicators. Doran's method allows the analyst to estimate the relative hierarchical position of each state in a defined system or set of states, the rapidity of each state's rise and decline in relative power and to forecast the likely future for each state in the system under review. The mapped "curve of relative power" for a state also allows the analyst to determine "critical points" on the curve. These points usually correlate strongly and positively with the incidence of conflict initiated or encountered by the state in question. 19 The Gouldian approach of time's cycle and time's arrow stands apart from the cyclic theories of international relations due to its non-systematic nature. It should be seen more as a heuristic metaphor without any direct applicability than a specific explanatory theory. However, the Gouldian approach may bring benefits to the analysis of intra-NATO disputes in three ways: First, the Gouldian approach allows for analysis of relatively specific and narrow themes. NATO's burden-sharing disputes do not represent such grand-scale developments as many of the above contributions on power cycles or transitions do. Instead, they, at least until present times, have been relatively limited in scope and taken place within a given institutional framework. The issues at stake do not deal with grand scale studies of states' rise and fallto bring the idea to a head, we are not talking about a Doranian rise or decline of U.S. relative power when discussing for instance the expressed American dissatisfaction regarding, for instance, alleged Canadian defence under-spending. Burden-sharing disputes operate below the radar of grand scale cyclic theories of international relations, which is why alternative approaches are needed for their analysis.
Secondly, this open-endedness of the approach enables a nuanced account of long-lasting issues in international security; it helps us to distinguish elements of continuity and change within a given issue-area. In other words, the application of Gould's arrow and cycle approach enables discovering typological regularities in the disputes. These regularities can spur the search for underlying theoretical explanations or forecasts, which can in turn be tested through within-case analysis. 20 A third feature of the Gouldian approach which shares similarity with Doran's is the propensity for forecasting intra-NATO disputes. As previously argued, whereas time's arrow captures the unprecedented uniqueness and distinctive character of sequential events, time's cycle provides these events with another kind of meaning by evoking lawfulness and predictability. 21 Thus, the domain of cyclic provides a stronger basis for forecasting than arrow.
To clarify this idea, a few lines on forecasting may is in order. A well-known challenge to any forecasting is the question of uncertainty, as all the information pertaining to the future is uncertain. While there can be many sources of uncertainty, it is convenient to categorise the character of uncertainties as either aleatory or epistemic. 22 An aleatory uncertainty is one that is presumed to be the intrinsic randomness of a phenomenon. Such uncertainty is characteristic to many issues of strategy and international relations uncertainty is inherent in the matter itself and it cannot be mitigated by collecting new information. 23 For instance, even the best forecasts are simply unable to take into account all the randomness and social indeterminacy affecting NATO members' policy-making. This aleatory uncertainty aligns among other things with such political events as unexpected election results, sudden death of key policy decision-makers or major changes in member states' domestic political scene.
Then again, an epistemic uncertainty is one that is presumed as being caused by the lack of knowledge, or data, but uncertainty is not inherent in matter itself. Examples of issues involving epistemic uncertainty might be certain societal trends such as demographic development in NATO members, or the available weapons technologies in the Euroatlantic region on a given time scale: basically, both developments can be forecasted with a relatively good probability. Forecasting tends to involve both types of uncertainties, even though it may be difficult to determine whether a particular uncertainty should be put in the aleatory category or the epistemic category. 24 It seems that both time's cycle and time's arrow involve some extent of epistemic and aleatory uncertaintyboth contain some predictable and some random elements. Yet, following Gould's argumentation, time's cycle tends to align more with epistemic uncertainty whereas time's arrow involves more aleatory logic. NATO's burden-sharing disputes tend to deal with a relatively limited number of issues, most of which tend to repeat over time. This is something over which at least a rough forecast, based on our existing epistemic knowledge, can be made. Yet, it should be noted that NATO burden-sharing agenda also evolves over time. This "direction," or the arrow of the disputes is, however, more difficult to estimate as a number of (aleatory) random elements have a big influence on it.
To sum up the "promise" of the proposed analytical setting: the concepts of time's arrow and time's cycle enable us to discover elements of repetition and elements of change in NATO's burden-sharing disputes. This kind of approach in turn allows a nuanced reading of the issues at hand and serves in assessing limited and mid-range phenomena in international security. Moreover, the approach gives some basis for forecasting, even though it cannot be counted as a theory. However, before diving into these questions, it is time to have a closer scrutiny of the burden-sharing disputes themselves.

Research material: post-Cold War NATO burden-sharing disputes in scholarly literature
Turning now to the empirical side of the article, this sub-section seeks to unpack the entity of NATO's burden-sharing disputes and categorise the individual cases into identifiable classes. Such a more detailed breakdown of disputes enables their assessment in the context of cycles and arrow(s), and thereby helps to illustrate the possible added value of Gouldian approach as a conceptual tool. The material here lists the diverse range of topics since the early 1990s, which has been portrayed in the academic literature as burden-sharing disputes within the Alliance. The collection of analysed literature was based on the following principles: it was conducted between December 2018 and May 2019 by utilising EBSCO Discovery Service database and supported by free internet search. 25 The search term used was "NATO burden-sharing." While the search yielded several hundreds of results, the main attention was given to texts describing and analysing past or current burden-sharing disputes within the Alliance. Therefore, in the search the relevance of each text was determined on the basis of whether it helped to identify additional NATO burden-sharing disputes, less interest was given to what the papers had to say about the issue itself. This way, the search yielded roughly 20 burdensharing related disputes as discussed in books, peer-referee journal articles, and shorter policy brief type of papers. The texts were thus selected on the basis of their subject matter, not any given publication forum. After that, the selected material was subjected to qualitative content analysis to identify the source(s), or topic(s), under dispute.
The most obvious research material, the political expressions by NATO member state leaders (speeches, summit documents) was omitted because of its contingent and selective nature: in their public messaging, politicians tend to focus on the most pressing issues of the day and to repeat their key messages, instead of seeking to sketch a larger picture. In contrast, scholarly literature as a writing format favours turning the attention also to less dominant, even neglected, themes. Likewise, policy briefs, which may not otherwise yield the academic standards of genuine research, have often written with the goal of raising awareness of less discussed but potentially relevant themes, which is why they were included to the research material.
To facilitate familiarisation, the selected material is classified below based on three categories or issue-areas: (1) burden-sharing disputes related to changes in NATO's external security political environment, (2) disputes related to NATO military operations or operations by NATO members (i.e. the US) outside formal NATO framework, and (3), disputes involving internal developments in member countries and institutional changes within NATO with a burden-sharing dimension. Each dispute or driver of disputes is followed in endnotes by sample source(s) in which it has been discussed. Thus, the literature review exposed 20 NATO burden-sharing disputes during the post-Cold War decades, ranging from grand strategic-level themes covering whole NATO, to specific policy questions relating to particular member countries, as well as from enduring questions familiar from NATO's history to more time-specific ones.
The above categorisation is not flawless and cannot be exhaustive, however. To begin with, the issues discussed in the empirical material are not commensurate. Some of the issues highlighted here represent the content of the debates, while some other are rather more underlying conditions that aggravate, or intensify, burden-sharing debates. Secondly, the discovered 20 disputes only reflect views presented by academics, leaving policy disputes below the radar of publicity into shadow. The history of NATO provides us with examples of numerous challenging bi-or multilateral negotiations, or hard talks at the North Atlantic Council which are not publicised at the time. Thirdly, the material here represents issues that have been highlighted in scholarly texts: utterances of decision-makers on burden-sharing are not included for reasons discussed above. Fourthly, the categorisation may contain partially overlapping themes and, as said, it only represents the debates taking place since the early 1990s. Finally, what is counted as a dispute is always a subjective undertaking, as it is the case with the precise location of some disputes within this categorisation: both understandings are open to alternative readings.
These benefits and limitations notwithstanding, the review should highlight the multidimensional nature of NATO burden-sharing, the diverse ways it can be perceived, and the wide range of topics that can lead to burden-sharing disagreement within NATO. What is at stake is an evolving issue, which seems to be open to changes in regional security environment, organisational dynamics, inter-state bargaining, beliefs, or perceptions.

Cycles of burden-sharing disputes
While the previous categorisation may help us to sort out post-Cold War burden-sharing disputes into a certain rudimentary classification, its usefulness in delineating the future nature of the issue at hand is limited. What would be more valuable and more policy relevant would be to sort out variables or factors that have led to burden-sharing disputes or driven them into a given direction; in other words, the conditions leading to disputes based on past experience. Accordingly, this section seeks to observe the above disputes in the context of time's cycle and time's arrow; that is, how can the debates be located into the framework of repetition, lawfulness, and predictability on the one hand and direction, uniqueness, and distinctive character of sequential events on the other.
As one studies the above categorisation through the lens of metatheoretical stances discussed earlier in this paper, the setting can be arranged into two main clusters of disputes with each having two sub-clusters. On the one hand, we can construct a realisminspired "axis of external security," emphasising NATO countries' search for stability and collective security and relating to NATO's external threat environment (either related to Russia or NATO's out-of-area operations in third countries) and institutionalisminspired "axis of internal unity," dealing with issues involving NATO's internal dynamics (either U.S. domestic issues or questions related to European members) on the other.
Starting with the "axis of external security," the four sub-clusters of burden-sharing disputes can be characterised as follows: First, large-scale geopolitical changes, either contractions, or expansionist, moves related to Russia have tended to raise challenges related to burden-sharing to the transatlantic agenda. In fact, the transformation of Russia conceived with the collapse of the Soviet Union can be said to constitute the start of the post-Cold War era. Historically, changes in Russia's position have led either to NATO's expansion, or to added need to support exposed member countries. Both are, in turn, apt to lead to intra-NATO discussion on how to manage and finance such changes. Russia's central role is of course understandable given NATO's traditional role as a pact dedicated to contain Russia's possible aggression in Europe. In the case of Russian contraction, NATO enlargement has led to additional costs from guarding NATO's expanded northern and southern flanks. In a more general sense, the costs of NATO's expansion embrace infrastructure in the new member-states, modernisation of their armed forces, enhanced reinforcement capabilities, the thinning of forces to defend longer borders and larger areas, and the increasing problems of decision-making in a larger NATO. 46 Then again, Russia's expansive moves in Ukraine have led to the need to safeguard exposed eastern member countries, leading to calls for vigilance among Alliance members and to related burden-sharing disputes. It is noteworthy that other major geopolitical post-Cold War transformations, in particular the Arab Spring, seem to have had only indirect impact on the internal NATO debate with the exception of Turkey's dissatisfaction regarding Alliance support during the Syrian civil war. 47 In any case, Russia-related developments seem to materialise in the political space of NATO's burden-sharing disputes in the form of questions on how to accommodate the costs related to NATO enlargement or how to share responsibility of exposed NATO members.
Second, a correlation seems to exist between burden-sharing disputes and military operations, either by NATO or by US-led coalitions. The debate has taken many forms, including questions such as who pays, who should participate, or how participation should be organised, be it NATO military operations in Yugoslavia (1994/5 and 1999), Afghanistan (2002-2014), or Libya (2011). Also, major US-led interventions in Iraq outside the NATO framework (1990/91 and 2003-2014) led to disputes within NATO as well. Then again, based on the studied material, this was less so in relation to military operations against ISIL in Syria and Iraq. Likewise, smaller scale NATO crisis response operations, such as those taking place on the Mediterranean, or off the Horn of Africa, do not seem to have led to similar disputes. Problems associated with out-of-area operations were absent from NATO's agenda during the Cold War decades, as the Alliance did not engage in the actual use of force. Thus, it seems that provision for crisis management measures, which for the first time appeared in the Alliance's 1991 Strategic Concept, has amplified burden-sharing disputes by NATO nations. Regarding the coming years, much will of course depend on whether NATO will continue to focus on collective defence and the territorial integrity of its members, as has strongly been the case since 2014, or whether out-of-area operations will again gain more prominence. The types of burden-sharing disputes aggravated in this context deal with questions of who does what in out-of-area operations, who takes risks, and who eschews them by employing 'caveats' (i.e. restricted command authorities granted to NATO commanders).
A third correlation relates to burden-sharing debates and phases of US retrenchment and renewal with respect to its central leadership role in the Alliance. This source of burdensharing tension dates thus to the early phases of NATO. When Washington has sought a retrenchment from the world, it has traditionally increased pressure on Europe to do more. Then again, during times of increased foreign policy ambition, the USA turned first to its traditional leadership role in the Atlantic Alliance. 48 Examples of major phases of US retrenchment include the early years of the Eisenhower Administration ca. 1953-54; the early 1970s, as well as in many respects the times of the Obama and Trump Administrations. 49 Another variation of the same theme have been the occasional phases of elevated US interests to other parts of the globe, in particular East Asia (Obama's pivot to Asia or "rebalance" towards Asia policy), or the Middle East (Operation Iraqi Freedom during the Bush Jr Administration), all followed by renewed debate on responsibilities and share of burden within NATO. Thus, the specific burden-sharing tension related to US involvement relates above all to overall Alliance defence expenditure.
Finally, the debate has on a number of occasions circled around the proper level of commitment of non-US, particularly European, NATO members. This cluster of disputes has involved a question of possible growth in European strategic autonomy above all in the light of the emergence and growth of the EU's Common Security and Defence Policy (CSDP), the relative weight of military vs. non-military contributions to Alliance collective security; 50 or in a wider sense, diverging American vs. "European" strategic cultures. Disputes on these questions are roughly as old as NATO itself, but they have gained more prominence during the post-Cold War era. They point to wider cultural divides within the Alliance and suggest a relatively permanent undercurrent in burden-sharing disputes. It is noteworthy that whilst most criticism towards European contributions relates to its perceived insufficient efforts, European NATO members are occasionally also criticised for excessive search for autonomy within the framework of the EU's CSDPin a way, for doing too much, or in any case, the wrong things. The occasional US inability or unwillingness to see the EU as a political union with a nascent defence dimension of its own may be one underlying factor here. This problemacy disembarks on intra-NATO question of the relative weight of military vs. non-military contributions to collective security. Following these thoughts, Table 1 below seeks to arrange the cases presented above according to factors that appear as cyclic drivers for intra-NATO disputes. It can be argued that in the light of NATO's history since the early 1990s, burden-sharing has turned into a political dispute among member countries in four relatively repeating and potentially concurrent categories as shown below.
As said, these four driversgeopolitical change related to Russia, significant NATO out-of-area operations, US retrenchment or renewal, or European passivity or activism do not appear in turns or in any specific order, but may appear in political space simultaneously so that NATO often faces several burden-sharing drivers and disputes at the same time. Likewise, the frequency of their appearance in political space is governed by political considerations in member states. Yet, what is more relevant for our purposes here is that they together suggest the Gouldian concept of predictability and lawfulness.

Arrow of burden-sharing disputes
Based on the above empirical material, how can we discern a "direction" or the time's arrow of burden-sharing disputes?
To begin with, the studied literature provides cases of burden-sharing disputes, which do not fit easily, if at all, in any of the above cyclic categories. First, as Hartley and Sandler point out, the advent of new (weapons) technology (item 1c in the category above) may also have burden-sharing effects. For instance, to develop high-technology defences requires massive research and development budgets and large investments in weapon systems, but all NATO countries are not equal in terms of resources to do so. On the other hand, once a technological breakthrough is achieved in a given domain of technology, the discovery can be applied to enhance the weaponry of other allies, thus providing non-rival, but excludable benefits with stealth technology being one example. 51 The lacking references to weapons technology in the literature might initially appear as somewhat surprising. However, whilst the development of technology may have often had burden-sharing effects, its role seems to be too subtle, and possibly infrequent, in order to rise directly to NATO's political space. After all, technology is a domain that is relatively unfamiliar to most politicians and government officials. That may be a reason why technological development tends to raise to political space only superficially and on very general level. Another branch of burden-sharing disputes suggested in the studied literature is apparently caused by the increased difficulty in sustaining NATO's The increased difficulty in sustaining NATO's "value narrative" during the Trump Administration The relative weight of military vs. nonmilitary contributions to collective security "value narrative" and the rise of populism and/or illiberal democracies in certain member countries (item 3i above). Such social phenomena are difficult to locate in a cyclic framework as they appear to have emerged to NATO related scholarly attention only during the last few years. 52 Together, these two sources of NATO burden-sharing disputes highlighted in the literature can be regarded as being too subtle, or too recent, to be assessed in terms of cyclic pattern. Instead, technological change and changes in members' domestic political scene can be interpreted as indications suggesting an arrow kind linear progressionan ongoing uni-directional change within the Alliance with potential effects on intra-Alliance relations. More explicitly, these indications of uni-directional change can be positioned in a longer continuum of NATO's development; that is, the expansion of the scope of burden-sharing debate during the post-Cold War decades. Throughout the Cold War years, NATO burden-sharing was predominantly measured in terms of cost sharing, that is percentage of GDP spent on defence. At that time, the focus was on the input side of defence, or how much each NATO country contributed. 53 The end of the Cold War led to changes in NATO's burden-sharing agenda both externally and internally. NATO's external security environment became more permissive due to the collapse of the Warsaw Pact and the Soviet Union. Later on, the emergence of the European Union's CSDP since the late 1990s added its own twist to the transatlantic agenda. One tangible element was the new post-Cold War focus on the ability and political will to project military power far away for a long time and to emphasise deployability and sustainability. Spearheaded by the United States, NATO approved various schemes to convert territorial defence forces into expeditionary forces: the Defence Capabilities Initiative from 1999 and the Prague Capabilities Commitment agreed to at NATO's summit in Prague, November 2002, are both examples of attempts to transform the European armed forces to meet new security threats beyond the continent. 54 On the other hand, the end of the Cold War and the emergence of the Global War on Terror since 2001 changed also the internal NATO landscape of burden-sharing. To begin with, a new emphasis on out-of-area operations, crisis management, and broader conceptions of security came to dominate discussion within the Alliance since the end of the Cold War. 55 Accompanied to these developments was a vigorous discussion of the relative weight of input versus output measures in the Alliance context: was NATO about quantitative defence spending, or should it emphasise more the effectiveness of spending, risk sharing, superiority of training and equipment? 56 These new understandings on burden-sharing were demonstrated in NATO summits in Istanbul (2004) and Riga (2006). The developments since the 2014 occupation of Crimea have led to reaffirmed emphasis of "traditional" input-oriented burden-sharing discourse within NATO with the 2 per cent GDP goal agreed, as the Wales September 2014 Summit declaration manifests. However, whilst the Ukrainian crisis has manifested that inputs still matter, NATO's shift to qualitative indicators have not lost their significance. 57 What has resulted is an expanding political space of Alliance burden-sharing. In fact, it seems that it has expanded from cost-sharing into a number of inter-related, but partly incompatible and image-driven discourses in the transatlantic community and continues to do so. Indeed, it is becoming more and more challenging to define the precise boundaries for issues related to burden-sharing. In the words of Ringsmose, as NATO has transformed itself from a collective defence organisation to a projector of stability, so have the parametres and vocabulary of the debate transmogrified. 58 Today, as the findings of the research material manifest, debate on burden-sharing covers diverse questions. These include, but are not limited to, such issues as what is entailed in being a "good" Alliance member, or how a "burden-shifter" should be defined; what is the weight of other factors and forms of contribution than defence spending, like willingness to risk one's own troops, or to accept casualties for common good during operations; what should be the weight of the availability of deployable forces for NATO missions; or how one should assess a constructive approach in Alliance decision-making bodies in NATO context, or generally reputational resources of given member states. 59 In a wider sense, "burden-sharing" in the post-Crimea context, is increasingly about fundamental questions on how security, threats and the appropriate ways to address these threats are interpreted.

Expect a spill over
The "arrow" of this course of developments suggests that the scope and diversity of NATO burden-sharing disputes is growing and likely to grow further in the near future. In effect, the burden-sharing discourse is in the process of spilling over from its traditional domains. In terms of forecasting, this course of development suggests that this expansion will take at least three forms in the next few years: First, domestic politics within members seem to count more and more in the development of burden-sharing disputes. This dimension is not new, as the age-old guns vs butter predicament has always loomed large among NATO members, and as the US Congress has often been the venue for such arguments. Still, in most NATO countries, questions of security policy have traditionally been left outside of day-to-day domestic political quibbles. However, the advent of the rise of populism, non-liberal political forces in several NATO countries, and the overall increasingly transactional approach to international relations, are likely to shape the domestic willingness in several member countries to engage in multilateral security co-operation.
Second, whilst the "classical" debate on military expenditure is likely to remain, or even dominate NATO agenda, the transatlantic discourse on burden-sharing is increasingly likely to cover non-military domains such as comprehensive security, resilience, security co-operation, and diplomacy. Instead of closed-door interstate bargaining, the political space of NATO seems to be more and more dominated by a bidding race both on NATO and domestic fora of what is one's contribution to common good.
Third, burden-sharing disputes are likely to extend beyond the level of NATO itself as partner countries are also likely to be dragged into these disputes and into the discourse of burden-sharing. Be it defence co-operation, joint exercises, or participation in peace operations, partners are likely to face an environment emphasising demands of good partnership, reputational resources, constructive approach, and other forms of contribution from the involved countries.
These developments notwithstanding, the intensity of burden-sharing disputes is not likely to endanger the existence of the Alliance. During its 70 years' existence, NATO has proved its ability to withstand internal strains. Even though many contributions, such as Richter and Driver have regarded unequal burden-sharing as a threat for NATO 60 ; and this article started by noting the inherent dangers caused by burden-sharing disputes, the overall picture seems to be more multi-dimensional. After all, in many respects the Alliance indeed faces change and progress though argument. Debate may be one of NATO's defining characteristics. It may also be necessary in order to define and redefine its strategic purpose. In this regard, disputes related to burden-sharing can also be seen as a manifestation of the Alliance's ability to withstand and accommodate internal disagreement, sometimes for protracted periods of time.

Conclusions
As an outcome of the diverse interplay of realist and institutionalist logic, NATO burdensharing disputes should be seen as a continuous struggle within a commonly agreed framework or a regulated struggle, the agenda of which is continuously evolving. This article has applied Stephen Jay Gould's notions of time's arrow and time's cycle to capture this logic and sought to produce a few hypotheses based on these notions. The Gouldian approach of time's cycle and time's arrow is above all a heuristic metaphor, without any direct applicability. Indeed, it stands apart from the cyclic theories of international relations due to its open-endedness and non-systematic nature. However, even though it lacks rigorousness, the Gouldian approach opens a promising way for an insightful and nuanced account of a relatively narrow and long-lasting theme of international security such as NATO's burden-sharing disputes. It helps us to discover how some things tend to change while the debate in other respects may be primarily repetitive.
To be more precise, the proposed approach facilitates the analysis of NATO's burdensharing disputes in two ways. First, the application of time's cycle and time's arrow enables an insightful and nuanced reading of the disputes as an evolving historical phenomenon. Based on the approach, this article has identified four repetitive cyclic drivers for intra-NATO disputesgeopolitical change related to Russia, significant NATO out-of-area operations, US retrenchment or renewal and European passivity, or activism. Each of these repetitive and potentially concurrent drivers has its specific social and political logic in influencing the direction and tone of burden-sharing disputes.
These repetitive cycles of disputes co-exist with an arrow of a course of development, seeking to capture the unprecedented uniqueness and distinctive character of sequential events extending to several decades. In the context NATO burden-sharing, this article has suggested an expanding scope of disputes during the post-Cold War era as the most important element of time's arrow. Whilst the studied material indicates decades-long continuities in the transatlantic debate, such as the "traditional" NATO question of cost-sharing in the form of European vs. American contribution in the transatlantic security, the last 30 years have seen the emergence of new issue-areas of intra-NATO disagreement, the most notable being the response to geopolitical shifts and NATO's new role in out-of-area operations. In this respect, the findings of this study support arguments by Ringsmose and Jakobsen: we seem to witness a widening agenda of burden-sharing within NATO. 61 New issues and new lexica have come to supplement the post-Cold War burden-sharing agenda and there are indications that the scope of the debate continues to spill over to other fields. Donald Trump's ascent into power may have played a substantial role in this development, as well as other nationalist tendencies in a several NATO countries during the 2010s on the one hand. On the other, the widening political space of Alliance burden-sharing may have been accompanied with the overall expanding understanding of security and the Alliance's role in it.
A second type of value-added of the approach is more an emerging prospect than actual finding. It relates to the possibility of forecasting future burden-sharing disputes. Here, the starting point is drawn from Gould's argumentation has been that some burden-sharing issues are more likely to recur than others, providing different kinds of starting points for their forecasting. In Gouldian terms, the time's arrow captures the unprecedented uniqueness and distinctive character of sequential events, whereas the time's cycle evokes lawfulness and predictability. 62 However, here this article's efforts must stop as to elicit predictability does not mean to produce an explanatory theory or a forecast.
Indeed, the findings of this article should not be seen as the final word on the topic but rather as a starting point and as an invitation to critique, testing and further research through future within-case analysis. Are the suggested cyclic drivers and the suggested direction of burden-sharing controversies qualified and relevant? How to better describe the overall logic of cyclinessfor example, how often do certain kinds of disputes repeat? Would it be possible to estimate their frequency based on past experience? How, more precisely and more ambitiously, could it be possible to identify early warning mechanism of a looming burden-sharing dispute before it rises to the political agenda?

Notes Notes on contributor
Dr. Tommi Koivula works as an assistant professor at the Department of Warfare, the Finnish National Defence University. He defended his PhD in International Relations at the University of Tampere in 2004. Previously, he has been a lecturer at the University of Tampere and a visiting researcher at the University of Kent at Canterbury (UK) in 2001. Tommi's current research interests include NATO, EU security and defence policy and International Relations theory. His most recent major publications include The European Union and the Use of Military Force: Uncovering the Myths (Routledge, 2016) and Arms Control in Europe: Regimes, Trends and Threats (edited with Katariina Simonen, Finnish National Defence University, 2017). Currently, he is engaged in a research project on the past, present and future of NATO burden-sharing.