Understanding American Power : Conceptual Clarity , Strategic Priorities , and the Decline Debate

What does it mean for the United States to be powerful? The prospect of a decline in American power, especially relative to a rising China, has attracted considerable scholarly and political attention. Despite a wealth of data, disagreements persist regarding both the likely trajectory of the US-China balance and the most effective strategy for preserving America’s advantage into the future. This article locates the source of these enduring disputes in fundamental conceptual differences over the meaning of power itself. We map the distinct tracks of argument within the decline debate, showing that competing positions are often rooted in differences of focus rather than disputes over fact. Most fundamental is a divide between analyses dedicated to national capabilities, and others that emphasise mechanisms of relational power. This divide underpins how strategists think about the goal of preserving or extending American power. We therefore construct a typology of competing understandings of what it means for America to be powerful, to show that a strategy suited to bolstering American power according to one definition of that goal may not support, and may even undermine, American power understood in other ways. Policy Implications • Policymakers should be cognisant that the ‘power shift’ debate is not reducible to simple differences in forecasts regarding material capabilities. It also encompasses fundamental differences over what it means for the United States to be powerful. • When evaluating claims regarding the future balance of national capabilities, insist on clarity as to what of type of argument is being made. Does it: o Forecast a trend in conventional metrics, such as GDP or military spending? o Dispute the adequacy of these metrics to capture qualitative differences? o Predict that a specific policy intervention will succeed/fail in altering present trajectories? o Posit a critical weakness in economic, military or political infrastructure that could bring about a collapse in the capabilities of the US or its rivals? • When developing strategy, prioritise between alternative understandings of power. Which is more important: o The size of the gap in material capabilities between the United States and its nearest rival? o The United States’ ability to dominate others coercively? o The United States having sufficient influence to obtain preferred outcomes consensually? o The entrenchment and longevity of a world order with liberal characteristics? • Strategists should acknowledge that while sometimes the policies best fitted to pursuing these different ends may be complementary, often they are not. Necessary trade-offs therefore ought to be identified and evaluated in a deliberate, calculated and strategic manner. When was the last time you’ve seen our country win at anything? We don’t win anymore. . . No matter what it is, we don’t seem to have it. -Donald Trump, 18 April 2015 (Reuters, 2015). Global Policy (2019) 10:1 doi: 10.1111/1758-5899.12609 © 2018 The Authors. Global Policy published by Durham University and John Wiley & Sons Ltd. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. Global Policy Volume 10 . Issue 1 . February 2019 5

• When evaluating claims regarding the future balance of national capabilities, insist on clarity as to what of type of argument is being made. Does it: o Forecast a trend in conventional metrics, such as GDP or military spending? o Dispute the adequacy of these metrics to capture qualitative differences? o Predict that a specific policy intervention will succeed/fail in altering present trajectories? o Posit a critical weakness in economic, military or political infrastructure that could bring about a collapse in the capabilities of the US or its rivals?
• When developing strategy, prioritise between alternative understandings of power. Which is more important: o The size of the gap in material capabilities between the United States and its nearest rival? o The United States' ability to dominate others coercively? o The United States having sufficient influence to obtain preferred outcomes consensually? o The entrenchment and longevity of a world order with liberal characteristics?
• Strategists should acknowledge that while sometimes the policies best fitted to pursuing these different ends may be complementary, often they are not. Necessary trade-offs therefore ought to be identified and evaluated in a deliberate, calculated and strategic manner.
All the talk of America's economic decline is political hot air . . . so is all the rhetoric you hear about our enemies getting stronger and America getting weaker. Let me tell you something. The United States of America is the most powerful nation on Earth. Period. It's not even close.
-Barack Obama, 13 January 2016(White House, 2016 The complexity of the decline debate The standing of the United States in the global distribution of power, and its trajectory, is not a subject that has struggled for attention. The suggestion that relative decline might be underway gives rise to strong and divergent reactions among International Relations scholars (Brooks and Wohlforth, 2016a, b;Ikenberry, 2011;Layne, 2012;Mearsheimer, 2001), news media (Bryant, 2015;O'Reilly, 2015), and the public at large (Wike et al., 2015). The questions of whether US power is waning and, if so, what should be done to forestall or reverse this have been the focus of heated disagreement at the highest level of electoral politics. We argue here that these enduring controversies cannot be reduced simply to differing assessments regarding trends in empirical data, that is, to narrow disagreement about the present and future balance of national capabilities. Rather, much of the decline debate is rooted in profound underlying differences regarding the nature of power in international relations, that is, about what it means for America to be powerful. In addition, this debate often gives rise to disagreements over what national strategy is best suited to preserving and extending American power. These disagreements, we argue, are not limited to the effectiveness of specific policies in advancing an objective that all share. Rather, they encompass fundamental divergences over the nature of the objective itself.
This article seeks to make two contributions, each aimed at enhancing the conceptual precision with which debates regarding American relative power are conducted. First, we offer a clarifying framework that systematises the underlying structure of disputes over whether a shift in US relative power should be anticipated. Second, we propose a typology for categorising national strategies that aim to preserve or extend that power.
We begin by identifying four distinct tracks of argument within those debates about the US-China balance where power is equated with national capabilities (see Figure 1). These are: differing predictions regarding trends in conventional metrics of national capability; claims that these metrics are insufficient to capture important qualitative differences between US capabilities and those of others; divergences of expectation as to whether policy interventions will succeed in altering present trajectories; and claims that the US or its rivals are vulnerable to a catastrophic sudden reduction in capabilities due to a critical weakness in economic, military or political infrastructure.
However, a different set of contributions to the decline debaterooted in a relational definition of powerreject the idea that power should be primarily identified with such capabilities. Turning to these, we identify four understandings of how power manifests that have featured prominently in arguments that contest the likelihood of American decline: the strong US position within global networks; structural power that affords the US privileged influence; soft power that augments US leadership by shaping preferences; and ideological hegemony, which subverts the autonomy of others, to American advantage.
Our purpose here is not to stake out a particular position regarding likely trends in US capabilities. Nor is it to argue that one conceptual approach to power, or one mechanism for its expression, is more 'correct' than others. Rather, our purpose is to clarify the underlying conceptual structure of the US decline debate, and to show that while differences over the interpretation and forecasting of empirical data certainly have their place within it, it also encompasses fundamental theoretical divergences regarding the nature of power itself.
Moreover, though it is often not articulated explicitly, much policy disagreementand associated political rhetoricrevolves around these basic conceptual differences over the meaning of power. Turning to national strategy, we present a typology of four alternative formulations of the goal of maintaining or extending American power. This is set out in Table 1. Each posits a distinct objective derived from its understanding of power, that is to seek to maximise one of the following: capability-gap; dominance; influence; and liberal order longevity. Crucially, the specific policies best suited to advancing each of these objectives are rarely identical, and often they are at odds. Not only is the academic decline debate sustained by conceptual discord about the nature of power, but competing policy approaches to preserving American power have their roots in disagreements around what it means for America to be powerful.
Defining, measuring and comparing relative power Robert Dahl provided political science with its classic definition of power: the ability of A to make B do what the latter would otherwise not. This relational conception of power, that is, power defined as a relation between actors, is not without its difficulties for analysts of the international balance of power (Baldwin, 2013). One is methodological: the existence of A's power cannot be verified unless exercised, and pending that its existence is open to doubt (Dahl, 1957). The other is conceptual: if we equate having power with getting one's way, how to make sense of our intuitive understanding that those we nevertheless persist in considering powerful regularly fail in their efforts to control others' behaviour, or are thwarted in the pursuit of their goals? 1 Likewise, a relatively weak state may achieve its objectives if it is astute in how it deploys its resources without this leading us to reclassify it as powerful. In other words: if we simply work backward from outcomes obtained to determine who is powerful, have we defined out of existence the conceptual space in which we evaluate the quality of strategies according to their success in translating power into the results desired by those who possess it (Waltz, 1979)?
A common move on the part of analysts attempting to overcome these challenges has been to focus on measuring the national material capabilities that states seek to leverage in their conduct of international politics. Understood this way, state power is defined as possession of a set of attributes produced by internal societal development (Organski, 1968). Measuring and comparing the relative power of states then becomes the task of quantifying and aggregating such attributes. 2 As Kenneth Waltz observed, 'states are not placed in the top rank because they excel in one way or another; they must score high across all relevant areas of capability, including size of population and territory, resource endowment, economic capability, military strength, political stability and competence.' (Waltz, 1979, p. 131). It is common for states to possess strength in one or more areas, but fall short in others, and thus fail to qualify as toprank powers. Such a state might, as in the case of several large European states today, be wealthy but lack a correspondingly high level of investment in armed forces. It might, like North Korea, have a large military establishment, but one grossly out of proportion to its economic foundation. It might, like Saudi Arabia, be endowed with valuable natural resources, but have only a modest population and inhospitable territory. Or, like Luxembourg, it might be wealthy but tiny: good for the bank balances of its residents, but fatal to any aspiration to geopolitical weight.
In this section, we do two things. In the first part, we map the tracks of argument within capability-focused efforts at measuring US power. In the second, we return to relational power, considering those approaches that focus directly on it and the mechanisms by which it operates.

The structure of arguments about relative capabilities
To illustrate the underlying structure of debates over the future of US power where this is understood as material capabilities, we focus on the US-China comparison. This is because China has stood out in making rapid, major advances across the range of great-power attributes, with the result that debates over American decline and China's rise have been symbiotic (Brooks and Wohlforth, 2016a). The argument map that follows is conveyed diagrammatically in Figure 1.

Understanding American Power
Track one: trends in conventional metrics A first track of debate centres on conflicting predictions of future trends in the conventional and most frequently-cited metrics of national capability. Most prominent in this category is GDP, where China's higher rate of growth over several decades has been the single greatest factor driving perception that a global power shift is underway (Quah, 2011). Disagreement in this area tends to focus on how long this can be sustained. It is universally accepted that Chinese growth will slow as the 'catch-up effect' takes hold (Matthews, 2006). Some go further, contending that ultimately all growth rates regress to the mean, which would make China overdue not merely for a gentle slowing but a major correction (Pritchett and Summers, 2013). The timing of any slowdown is of crucial importance because although China's GDP will soon match that of the US, its GDP per capitaa second key indicatorwill trail behind far longer, perhaps forever. 3 One study that extrapolates for China based on the 20th century growth patterns of Japan and South Korea foresees it attaining GDP per capita approximately half that of the US only by 2061, and its rate of advance tailing off thereafter. Such gains would represent spectacular progress, of course, but parity would remain far distant (Aiyar et al., 2013;Hoffman and Polk, 2014;Jiang and Yi, 2015). The conversion of economic base into operational military capabilitiesin which at present the US enjoys an immense global leadis core to any shift in the international balance, which makes military spending another ubiquitouslycited metric (SIPRI 2016). Accomplishing the conversion requires state capacity sufficient to extract the necessary resources for foreign policy purposes (Taliaferro, 2006), and sufficient political support for doing so (Kitchen, 2010), neither of which is automatic or immediate. Even where economic development ultimately is harnessed by growing state capacity, there may be a lag of decades (Zakaria, 1998). Further, even once military capabilities have been developed, power projection overseas may still be hampered by insufficient institutional coherence (Schweller, 2006), or by domestic political resistance (Dueck, 2006). China has increased its real-terms annual military spending by 132 per cent over the last decade and is committed to continuing increases. Forecasting whether this rate of increase will continue, and whether the US will meet it with increases of its own, is therefore central to determining expectations regarding power shifts. If present economic growth trends were to continue for an extended period, China could have the resources to equal the absolute dollar amount of US military spending even while a sizeable gap in per capita GDP persisted. 4 Demographic projections often loom large in differences of opinion regarding economic and military trajectories, since national population supplies the human capital required for production, the consumers to drive growth in domestic markets, and the personnel necessary for armed forces. The stark fact of China's enormous population -1.4 billion people to America's 323 million-goes a long way to explaining the widespread intuition that its rise presents a challenge to the US of a different order than that posed by economic rivals like Germany or Japan during their periods of rapid postwar growth. Equally, a common riposte to any prediction that China will eventually overtake the US is to note that the US has, for a developed country, a robust level of annual population growth. This is not, it should be acknowledged, primarily due to fertility rates, which are historically low, but to net immigration (Livingston, 2018). China, however, lags behind on both. This is not by chance: legally restricting reproduction was a deliberate step in China's strategy to accelerate development and raise living standards (Fong, 2016). But unless something changes, an unintended consequence will be a rising dependency ratio of non-working to working population that could block the country from reaching American levels of wealth: China, in the common phrase, may get old before it gets rich (Cai and Wang, 2006;Hilton, 2011). China could seek to mitigate the impact on government finances by declining to provide American levels of welfare for the old, but this would not erase the underlying economic drag of a reduced workingage population. 5 In the final analysis, arguments on this first track are reducible to predictive wagers regarding a fixed set of publicly accessible, widely-cited metrics. Disagreements of this type have the happy quality of being definitively resolvable through the passage of time and the accumulation of new statistical data. We may not be able to know for sure today who is (in)correct, but in due course we assuredly will.

Track two: deeper dimensions of superiority
The second track of debate centres on claims that the most frequently cited metrics fail to convey the true scale and substance of America's lead in capabilities. One such argument is that GDP statistics inadequately capture levels of knowledge-based productivity, innovative capacity and deep technological sophistication that render the US economy qualitatively superior to the Chinese (Brooks and Wohlforth, 2016a;Coyle, 2015). Another is that national GDP figures obscure how large a proportion of Chinese production is in fact owned and executed by corporate entities with a chain of ownership rooted in the United States (Starrs, 2013). Similarly, some assessments of relative military capability argue that raw spending flows are too crude a measure to capture differences in the extent and quality of actual capabilities. Close analysts emphasise that US superiority rests not merely on higher annual budgets today, but on the accumulated fruits of decades of investment in research and development, technology, doctrine and human resources (Beckley, 2011;Itzkowitz Shifrinson and Beckley, 2013).
This track of argumentthat predictions of decline rest on metrics that are too superficialcalls on participants in debates about power to show a more sophisticated understanding of the relationship between indicators of capabilities and capabilities themselves. It is a call, in essence, for more and better data. In the economic realm, we might perhaps monitor the number of transnational corporations in the hands of Chinese nationals. 6 On innovation capacity, analysts might incorporate data on patents, or the degree to which Chinese-owned intellectual property is exploited globally by native start-ups in a manner comparable to the great global success stories of US tech. 7 To detect true erosion of the United States' military superiority, it will be necessary to assess whether increases in Chinese spending have translated into shifts in effective force capacity (Heginbotham et al., 2015). The accumulation of new and more advanced hardware is one obvious manifestation of this. US defence planners adjusting their strategic plans to reflect the reality of new Chinese capabilities is another (Kazianis, 2015).
Within this second track, some debate may endure over which metrics of capability are most equivalent to the capabilities themselves. Nonetheless, there is nothing intrinsic within this track to prevent, in principle, broad consensus surrounding the trajectory of the US-China balance. The task of those who argue that prevailing metrics do not tell the whole story is to respond with an extended menu of metrics to facilitate the more comprehensive and accurate assessment of capabilities they wish to see.

Track three: betting on the outcome of policy interventions
The third track of debate, which branches from each of the first two, is focused on anticipating the success or failure of deliberate policy interventions. The first task of analysts of relative power is to assess current levels of national capability, and project how trends in these might unfold into the future. National governments, however, having engaged in such analysis themselves, are likely to take action seeking to alter the course of those trends. A key question at issue, then, is whether such policy efforts will succeed.
For example, all agree that much of China's economic growth has been thanks to adoption or imitation of Western technology and designs, and that ultimately it must increase its capacity for home-grown innovation (Abrami et al., 2014;Shah et al., 2015). The Chinese government, being acutely aware of this, has formulated policies intended to overcome this barrier to progress, but assessments of their likely impact differ (Huang, 2013;Lange, 2015;McGregor, 2010). Likewise, Beijing's defence budget increases are linked to a strategic plan for modernisation of China's armed forces (Cordesman et al., 2013;State Council Information Office 2015). Further, simply replicating present US military assets would not afford China true parity of capabilities with the US. Achieving that requires fostering the capacity to conceive and develop the as-yet-unknown weapons, defences and doctrines that will be decisive in the military engagements of the future (Cheung, 2014;Rogers, 2014). While a policy of cyber-espionage enabling imitation of US technological advances may be a partial workaround for inferior R&D capacity, most analysts agree that in the long run it is no substitute. 8 Demographic trajectories are also subject to policy intervention. China has already loosened population controls, though, as some have pointed out, policies to actively boost birth or immigration rates are politically challenging (Luo, 2015;Qiao, 2006).
We should remember, meanwhile, that the United States' advantages are not immune from being augmented or undermined by government policy interventions, so arguments on this track must also anticipate these. This is most obvious when it comes to choices regarding levels of technology and infrastructure investment. But to give an example less frequently considered as relating to the international balance of power: America's relatively optimistic demographic projections could be significantly altered if a new era of restrictionism in immigration policy, such as that advocated by President Trump and other senior officials since 2017, were to take hold (Baker, 2017).
Like the first two tracks of argument, it should be possible to establish over time, with detailed study, the impact of policy choices on relative capabilities. Because policy interventions are multiple, and dynamic in nature, any single assessment of their effects will necessarily be a snapshot of the situation at a given point in time. And further shifts in policy will require further updates to the projected trajectories of trends upon which their impact is felt. But even with these challenges stipulated, there is nothing intrinsic to this track of argument that precludes in principle the formation of broad consensus.
Track four: identifying critical vulnerabilities, or sources of fundamental disruption Fourth and finally, there are the prophecies of sudden unravelling. Within this track sit claims that one or other competing society or economy has a singular flaw, dormant for now, that will one day manifest in a crisis. This is asserted quite often with regard to the Chinese economy, with concern expressed about liabilities concealed in an opaque shadow banking system (Barboza, 2013;McArdle, 2013), a property bubble (Glaeser et al., 2016), and stock market distortions arising from government intervention (Lee, 2015). A major unwinding originating in such deep vulnerabilities could not merely bend the curve of longterm growth but wipe out years or decades of prior accumulation of national wealth. The highly advanced and diversified economy of the United States, meanwhile, is generally presumed to be more resilient. But as the financial crisis of 2008-09 exposed, it is not without vulnerability to systemlevel risk (Financial Crisis Enquiry Commission 2011).
Other variants on this track of argument posit that a sudden disruption of the military balance could be triggered by a game-changing technological leap. It is clear that command of information and data is becoming an ever-more valuable currency, and cyber assets are now critical to many kinetic military operations (Bonner, 2014). At present, this appears to buttress America's advantage because of its presumed lead in developing the requisite skills. But possible futures are readily conceivable in which China, rather than seeking to replicate the full extent of US hardware and systems, instead acquires the asymmetric means to neutralise Understanding American Power them. This could take the conventional form of new battlefield weaponry, such as the 'carrier-killer' missiles some forecast will radically impede American freedom to operate in the South China Sea (Myers, 2018). Or it might take less material form: offensive cyber capabilities able to disable the high-tech networks on which much of America's military and economic superiority rests (Department of Defense 2016). Related, there is the radical possibility that disruptive technology might profoundly change the nature of the capabilities that matter for the international balance (Mann, 1986). For example, some believe additive manufacturing (3D printing) will upend the prevailing trade order (McKinsey Global Institute 2013); or that a fourth industrial revolution of automation and connectivity is creating new categories of threat (UBS, 2016); or that innovation in power-generation will reduce the global economy's reliance on fossil fuels (Bradford, 2008). Such disruptions have the potential to shift what Mann calls the leading edge of power by fundamentally altering the strategic geography of the planet (Buzan and Lawson, 2015;Mann, 1986). All capability-centric projections must be appended with an asterisk accordingly.
Finally, perhaps the most common argument on this track centres on political risk. Sudden political change in China could have seismic economic consequences, a fact that gives even earnest liberals pause as to the desirability of rapid democratisation (Bell, 2015). The reverse is also true: if economic growth were to stall, and with it rises in living standards, the Chinese one-party system could collapse into disorder, or be forced to shore up its authority through escalating domestic repression (O'Mahoney and Wang, 2014;Shirk, 2009). US institutions, though of greater proven longevity, are not invulnerable. For over a decade, the American constitutional system has been put under escalating strain by polarisation and partisanship (Mann and Ornstein, 2016). The increasingly disorderly process of funding the federal government, with Congressional votes attended by threats of credit default and government shutdown, may suggest degrading functionality of key institutions (Weisman and Peters, 2013). The election of public office-holders who display overt antagonism toward liberal democratic norms has raised further concern over both the administrative competence of the government and the security of the rule of law (Frum, 2017;Klein, 2017;Sullivan, 2016). This has been exacerbated by criminal investigation of senior officials for collusion with Russian interference in the 2016 election, obstruction of justice, and financial impropriety. All predictions regarding America's future national power entail an implicit judgement on whether these present challenges can be contained or are indicative of a longer-term decay in the performance of American institutions (Ikenberry, 2002).
The significance of this fourth track of argument is to introduce uncertainty irrespective of the degree of consensus that might be reachable across the first three tracks. The occurrence of 'black swan' events that render the most assiduously constructed trend forecasts null and void is an occupational hazard of all that engage in these debates.
Nevertheless, these do not signify any internal contradiction that vitiates the purposefulness of forecasting itself.
In summary: mapping capabilities-focused arguments about the US-China balance reveals distinct logical tracks of disagreement that turn on identifying and prioritising data, forecasting trends, and judging the likely outcomes of policy interventions. Though this is a complex picture, claims within these arguments are commensurable, in that they are part of a single coherent enquiry. Analysis of capabilities, however, is only part of the debate around American power. Relational approacheswhich treat capabilities as relevant inputs, but more closely associate power with obtaining the outcomes one desires through interaction with otherspose a competing understanding of what it means to be powerful. We now turn to these.

Forms of relational power
Both the potential utility and the complexity of measuring material capabilities as a proxy for power will by now be apparent. Some analysts, however, noting the limitations of such metrics (Merritt and Zinnes, 1988), prefer to focus on the mechanisms by which states exercise it (Baldwin, 1989;Barnett and Duvall, 2005;Dahl, 1957;Frey, 1971;Lasswell and Kaplan, 2013). Capabilities-focused assessments of relative power implicitly assume a generally close correspondence between those capabilities and a state's ability to control the behaviour of others. By contrast, relational approaches posit that the link is more tenuous (Baldwin 2016). While not denying that capabilities are important, these approaches emphasise that 'power is not itself a resource', but rather 'resources are the media through which power is exercised' (Giddens, 1979, p. 91).
The primary importance of this move lies in its carving out intellectual space to argue that, in certain circumstances, a state's power might be maintained at the same level even if its ranking in material capabilities were to be diminished. In debates on American power, the mild version of this claim is that there is a lag: even if its advantage in capabilities erodes, the US will continue for a time to exercise disproportionate influence over others as a legacy benefit of relational pathways established during its period of material dominance. The stronger version of the claim would be that there is no necessary causal connection between one shift and the other: that the US can somehow indefinitely sustain a superior capacity to get its way that exceeds its comparative position in material capabilities.
We identify here four forms in which relational accounts suggest American power manifests: Network position: states do not try to influence others' actions exclusively through bilateral interactions at the top level of government. They operate within complex networks of connections, alliances, and understandings, which link both state apparatuses and civil society at multiple levels. Sometimes, these networks operate within international institutions and organisations; often they are less formal. They serve as delivery channels for coercive demands and persuasive efforts, and provide a setting for both to be responded to. Not all states are equal in this regard: the US occupies a central position within myriad networks, and some argue this affords America a valuable edge over rising competitors. As Anne-Marie Slaughter (2009, p. 94) succinctly states the case: 'In this world, the measure of power is connectedness'. If this is true, then the US might aspire, by retaining this advantage, to be no less powerful even as its margin of economic and military superiority eroded.
Structural power: the US is not just well-networked. It also enjoys a privileged position conferred by what Susan Strange (2015, p. 27) termed structural power: the ability 'to shape frameworks within which states relate to each other'. When a state has a great advantage in material capabilities, the mechanisms of structural power help convert this into influence over others (Gruber, 2000;Wade, 2013). Once established, however, such mechanisms can be persistent and resilient even if the underlying material superiority that gave rise to them erodes (Wade, 2011). Today's rules-based international order was established during a period of immense economic imbalance favouring the US, and the institutional legacy of that period continues to buttress US primacy (Gilpin, 2002;Ikenberry, 2001). An example of the United States' formal privilege is its special status and voting weight in institutions such as the UN Security Council, IMF and World Bank, which affords it substantial control over their agendas. 9 A less explicitly institutionalised, but no less real, instance is the 'exorbitant privilege' conferred by the US dollar's status as the dominant global reserve currency, which allows it to raise capital more cheaply and measurably boosts national GDP (Eichengreen, 2012;Norrlof, 2010). If the precedent of Britain's decline is any guide, there could be a lag of decades before such arrangements adjust to fully reflect any underlying shift in the United States' relative economic weight. 10 Soft power: states can also co-opt one another through the appeal of their own characteristics, something Joseph Nye captured in the concept of soft power. In the years since its initial coinage, 'soft power' has become a ubiquitous term, often deployed imprecisely as a synonym for all non-violent forms of influence, or even coercion by nonmilitary means. In its more specific original sense, however, it means displaying social, political and cultural attributes and values that others find attractive and seek to emulate (Gallarotti, 2010;Nye, 2002). Over time, it is posited, this fosters a commonality of worldview that leads others to 'want what you want' (Nye, 2004: 256). In policy terms, states may seek to operationalise their soft power through diplomatic means: framing agendas, deploying persuasive rhetoric and actively encouraging 'positive attraction' (Nye, 2011, p. 19). Manifested in this way, soft power can resemble structural and ideological power; most forms of preference-shaping overlap to some degree, and it is certainly true that soft power operates indirectlythrough what others find attractive rather than through negotiation of specific demands (Laifer and Kitchen, 2017). However, whereas structural forms of power operate at the level of decision-making and agenda-setting, and ideological power shapes unconscious frameworks, soft-power's focus on the specific dynamics of attraction and voluntary emulation distinguish its particular role.
The United States has an advantage in this regard, according to Nye and others, because its broadly liberal culture and political values have wider appeal across borders than those of rivals. This claim demands some cautionary caveats: that making soft power work as public policy is challenging (Laifer and Kitchen, 2017;Nye 2004); that the US regularly falls short of embodying its professed ideals; and that the world is far from unanimous and unqualified in its admiration of America (Wike et al., 2015). Despite China's alleged inherent disadvantages in soft power, the concept has attracted great strategic and practical policy interest there (Mingjiang, 2008;Wang, 2008;Nye and Jisi, 2009;Shambaugh, 2015). Nevertheless, in one 2016 ranking based on a compound metric of indicators, the US placed first globally for soft power, with China ranked just 28th (McClory, 2016). The extent to which soft power is ultimately derivative of material superiority (Kearn, 2011), or indeed may help foster it (Rose, 2016), is contested. But if the relation between the two were a simple one, we should expect a close correspondence between capabilities and soft-power ranking, and China's present position suggests at minimum a substantial lag before capability gains are translated into proportional growth in soft power.
Ideological hegemony: the deepest sources of influence over behaviour are ideational frameworks that shape preferences below the radar of explicit political consciousness. If a dominant state can embed its preferred ideas so deeply that others perceive them only as ambient assumptions about reality and appropriateness, this removes the need for coercive actions, or even persuasive ones, aimed at directing their behaviour. 11 From a critical perspective, this can be read as a form of hegemonic power: subversion of the autonomy of others by the exercise of invisible influence over the intellectual space in which they define their identity and interests (Cox, 1981(Cox, , 1983. One way of framing the history of post-Second World War US foreign policy is that it succeeded in doing exactly this: internationalising its own preferences by successfully establishing liberal norms as the universal standard of appropriateness for economic and diplomatic conduct. If this 'American system' proves capable of integrating and socialising rising powers, this might mean the sustained alignment of the norms of world order with US preferences long after present levels of US material superiority have past. 12 This type of power does necessitate a caveat, however. If what were once merely American preferences truly have been converted into norms widely regarded as possessing inherent legitimacy, is it accurate to characterise the continued hegemony of those norms as a manifestation of American power? If, hypothetically, the US does not wield liberal norms instrumentally, and does not consciously conceive of them as a vehicle for advancing specifically national objectives by manipulating others, thoughts, are they nonetheless best understood as tools of national power? Or, an even more acute question: if the US Government ceases to be a committed subscriber to liberal idealsdo they represent a Understanding American Power vehicle for American national power, or a constraint upon it? This question takes on pointed importance when we come to consider US strategic priorities, as we do in the next section. Just as in the case of economic, military and demographic capabilities, other governments can undertake deliberate policy efforts to gain ground on the United States in each of these four areas of relational advantage. How will we know if they are succeeding? In the economic sphere, China has been rapidly expanding its network of trade, finance, and investment ties via 'partnership diplomacy' (Zhongping and Jing, 2014). It has also created a major new formal institution, the Asia Infrastructure Investment Bank (AIIB) (Krauss and Bradsher, 2015;Pandey, 2015;Womack, 2013). The proportion of international lending that takes place via the AIIB and from China bilaterally will be important indicators (Ferdinand, 2016). Some shrinking of the overrepresentation of American nationals among top personnel at global economic institutions such as the IMF and World Bank might be a sign of reduced advantage. Even more so would be adjustment of the relative voting weights of the US and China in the governance of those institutions. The dollar's status as the world's primary reserve currency is a cornerstone of US structural power; a plausible proxy for the security of this would be the proportion of global reserves held in dollars, as well as its weighting in the IMF's basket of currencies (Bradsher, 2015). Beyond the economic sphere, a sign of erosion in the US position would be the loosening of strategic alliances in Europe and Asia. Even in the absence of formal treaty changes, shifts in rhetoric and resource allocation could be meaningful (Wright, 2017). It also would be significant if China were to become a leading originator of innovation in transnational governance; this has been the almost exclusive preserve of the US and its allies (Hale and Held, 2011). Consensus is elusive on what metrics best capture soft power, but as noted above, efforts have been made (McClory, 2016).
The deepest form of long-term influence over behaviour, ideological hegemony, is also the hardest in which to measure a shift. For China, to successfully challenge the status quo in this regard would be far more difficult even than matching US material capabilities (Beeson and Li, 2015). It would entail first articulating distinct principles of its own that run contrary to the established order, and advocating for them as a programmatic alternative to which it seeks converts. And even if this enjoyed a degree of success, merely carving out a Chinese sphere of ideological influence would not be sufficient to equal the accomplishment of the United States at peak power, which has been to establish its principles as the embedded normative architecture of a truly global order. Even by the most bullish assessments, this is a distant vision for China. 13 Nevertheless, informed reporting indicates that its leader's ultimate ambitions may be on this scale (Osnos, 2018).
National strategy and priority: what kind of power does America most want?

Four alternatives
Every US politician, official and strategist would say they wish America to have as much power in the world as possible. Such apparent consensus, however, obfuscates more than it reveals. Disagreements over how best to maximise American power are rarely reducible to purely technocratic differences over the effectiveness of a given set of policies. More often, they encompass deeper conceptual divergence as to the type of power the United States should prioritise possessing. Building on the distinctions established in the previous section, we propose here a typology of four distinct ways of thinking about power maximisation. These are set out in Table 1.
The first conceptual type is the capability-gap maximiser. To think this way is to equate an erosion of American superiority in capabilities, in and of itself, with a decline in power. Viewed through this lens, it makes no sense to speak of the United States managing to sustain its power advantage despite a shrinking marginal lead in economic and military capabilities. The capability gap is American power; loss Institution-building and diplomacy Example and socialisation of the former is ipso facto loss of the latter. Adopting this perspective simplifies the focus for any argument over the trajectory of US relative power (though as the relevant subsections earlier in this article make clear, this by no means makes all disagreement trivially easy to resolve). Strategists adopting this priority should not necessarily be dismissed as fetishising capabilities for their own sake; US primacy in capabilities may have beneficial consequences. Some realist scholarship, for instance, has argued that demonstrably unchallengeable American primacy in capabilities increases geopolitical stability because the only available response is acceptance (Wohlforth, 1999). The controversial 1992 draft Defense Planning Guidance produced by the US Department of Defense is an example of a strategy premised on prioritising capability-gap maximisation in stark terms (Gellman, 1992;New York Times 1992). The second type is the dominance maximiser. To think this way is to understand US power as the ability to compel compliance from other states in response to specific demands. Superiority in national capabilities is important to a strategist prioritising dominance, since success often depends on having the resources required to issue or imply credible threats. But this type of thinking differs from the previous one in its primary concern with relational outcomes: to satisfy a dominance-maximiser that the United States is powerful, US capabilities must be applied coercively, with positive results. Importantly, this mindset does not, as some others do, see implicit weakness in the resort to coercion, that is, as revealing that consensual preference-shaping must have been tried and failed. On the contrary: from a dominance perspective, the more disinclined others are to alter their behaviour, the greater the demonstration of power entailed in compelling them to do so. For this reason, dominance should not simply be viewed as the next-best, fallback approach of an unsuccessful influence maximiser. A strategist who attaches the highest value to dominance will therefore be chiefly concerned with, and identify the test of American power with, instances of open confrontation where another state defines its interests in opposition to those of the US and defies its demands. High-profile examples of such confrontations in recent years might be with the governments of Iran, North Korea, Syria or perhaps Russia. Milder forms of conflict, such as with allies, may also at times manifest similar dynamics. For a dominance maximiser, the measure of power is how frequently and fully the US can force others to back down once antagonistic positions have been adopted (Holmes, 2016).
The third type is the influence maximiser. This way of thinking shares with the dominance maximiser a concern with relational outcomes, that is, the ability to deliberately steer the behaviour of others. But it discounts the value of overt displays of coerced obedience. For a strategist prioritising influence, US power should be measured by how frequently and closely the outcomes of international interactions correlate with those for which the US has advocated, especially in collective decision-making settings. The primary form this advocacy takes is not explicit or even implied coercion, but rather the channels for diplomacy and preference-shaping highlighted by theorists of networks, institutions and structural power. The influence maximiser is distinguished from the dominance maximiser by the fact that coercive measures, though they can be deployed when necessary, are considered an unsatisfactory last resort, revelatory of weakness and symptomatic of the failure of the preferred currency of power. Importantly, the influence maximiser values the liberal framework of norms and institutions primarily for purely instrumental reasons: as a means by which the US can seek to orchestrate behaviour, at lower cost than coercion.
Establishing and maintaining influence through networks and institutions carries a price, however, in that the US must be willing to periodically forego opportunities for short-term transactional gain in order to be seen as upholding established rules. Since a liberal rules-based order is premised on the willingness of participants to prioritise absolute gains over relative, this entails a long-term trade-off that likely conflicts with the priority of the capability gap maximiser. Significantly, the influence maximiser does not consider that they are de-prioritising national power in this trade-off in order to pursue other benefits. The reason they are prepared to tolerate a smaller lead for the US in capabilities, if it contributes to retaining structures and institutions that privilege US influence, is that for them the latter is power of the most important kind.
The fourth and final conceptual type is the liberal order longevity maximiser. To think in this way is to identify American power with the established norms that constitute and underwrite the liberal rules-based world order. Unlike the influence maximiser, however, a strategist with this priority does not value liberal norms and institutions primarily as instruments for the United States to engineer specific outcomes favouring its concrete interests. Rather, they invest with overriding intrinsic value the objective of embedding the liberal order as a permanent foundation for international society. This makes it a goal of the highest priority to convince rising powers to become responsible rule-upholding stakeholders, even if accomplishing this requires both tolerating some reduction in the US lead in capabilities and trading away some of America's structural and institutional privileges.
This last entry is potentially the most contentious category within the typology. This is because it is a matter for dispute whether the establishment of liberal international order is correctly characterised as perpetuating specifically American power. The nature of ideological hegemony, successfully established, is that it limits the ideological bandwidth of actors across the system without their awareness (Buzan and Lawson, 2014). Even the powerful state whose past leaders originated the prevailing ideology may no longer perceive that prevalence as an exercise in power on their part. Certainly, those who object to the universalisation of liberal norms often view them as a manifestation of American power (Hardt and Negri, 2000). But if the liberal rules-based order survives and thrives as a de-territorialised hegemonic paradigm at the same time as US power in the preceding three sensescapabilities, dominance, influence wanes, this will assuredly be unsatisfactory to a great many American politicians and strategists, less willing to Understanding American Power equate the US national interest with the triumph of globalist liberal principles. In an inversion of dominance as defined earlier, those prioritising ideological power will consider it most deeply and successfully established when it is completely invisible.
Implications of the typologywhy does this matter?
The categories outlined above are ideal types. Actually existing leaders and strategists may shift over time in the type to which their thinking most closely corresponds. And an individual, or state, may at any given time manifest elements of more than one type in their statements and actions. The value of the typology lies not in its ability to seal all real-world actors neatly within boxes, but in its utility as an analytical tool to identify unresolved tensions within national strategy debates. The fact that individuals may vacillate unreflectively between the understandings of power distinguished in the typology serves to underscore, not undermine, its usefulness in this regard. Greater conceptual clarity around what participants in the debate about American power mean by power is crucial, because the four ways of thinking outlined here do not always have uniform or complementary implications for national strategy. Of course, some disputes do come down simply to disagreement as to the likely effectiveness of proposed courses of action. For example, individuals may differ over whether a particular economic policy will boost longterm growth; or a military investment programme will pay dividends in useful assets; or a particular package of threats and inducements will persuade an adversary to change its behaviour; or whether a negotiation strategy is likely to secure the desired outcome in an institutional setting. In such cases, there is a shared understanding of the objective, with divergence limited to the matter of what means will be most effective. Often, however, disputes that superficially appear to fit this description contain, concealed, deeper underlying divergences as to the kind of power that should be prioritised. That is to say, the disagreement concerns not merely appropriate means, but ultimate ends. This is significant because ways and means likely to advance American power understood in one way may not do so in others; indeed, they may actively undermine it.
For example, consider a strategy of major increases in US defence spending, combined with encircling military deployments in East Asia. Such policies might well maximise the capability gap the US enjoys over China; yet as a means of forcing Chinese compliance with US demands, that is, maximising dominance, its efficacy would be far less assured. At the same time, such aggressive cultivation and deployment of hard power resources might severely undermine US influence over China within international institutions across a range of issues. Or a second example: granting China higher status and increased voting weight in institutions of international economic and security governance. This might contribute to the longevity of the liberal world order by persuading China to become a responsible stakeholder; but it might at the same time reduce America's ability to use those institutions as vehicles for magnifying its own influence and engineering its preferred substantive outcomes. Or a third: continuing to trade with China on present terms. On balance, this likely contributes to preserving international structures that boost US influence, and to further embedding liberal order. It will also benefit the US economy in absolute terms. Over time, however, it is likely to facilitate relative gains for China in terms of its economic and military capabilities.
The point here is a simple one: while the headline goal of 'preserving American power' may be uncontroversial, different underpinning understandings of power often mandate markedly different courses of action in pursuit of it. Arriving at robust judgments about the policies, investments, and tradeoffs required for a successful national strategy demands conceptual clarity regarding the specific type of power to be prioritised. It remains, of course, intellectually possible to argue that one course of policy action optimally blends pursuit of all forms of power such that all are maximised at once. Indeed, this rhetorical move is a reflex for many in the political realm, since it avoids the need to overtly acknowledge uncomfortable trade-offs. However, the onus should be upon those who advocate for such a panacean position to articulate explicitly how their preferred course reconciles the quite distinct ends specified in the typology here.

Conclusions
The future trajectory of American relative power is a question to which a great many analysts and scholars have provided their own diverse answers. We have not sought to add yet another here. Instead, we have sought to clarify the nature of the debate itself, that it might be better understood. Why is this important? First, because arguments which seek to advance or rebut the proposition that American relative power is in decline often fail to render their conceptual underpinnings fully explicit. From the perspective of rigorous social scientific enquiry, there are intrinsic benefits to doing so. Lack of conceptual clarity often results in analysis that is partial, and debates in which participants speak at cross-purposes. Frequently, divergent conclusions about the future trajectory of American power hinge less on competing empirical claims than on what understanding of power is implicitly adopted at the outset. Today's decline debate is the latest manifestation of a conversation that has already run for several decades, the balance of mainstream opinion tipping back and forth with the headlines and political mood of the period (Cox, 2001). For contemporary arguments to amount to more than cyclical rehearsal of familiar positions, greater discipline regarding both definition of terms and tracks of argument is needed.
Second, the distinctions sharpened here are not just academic in their significance: they have implications for the strategies of leaders and the actions of states. That American leaders would wish to preserve or extend American power is a given. Effective strategy in service of that end, however, requires clear thinking about what it means for America to be powerful. Only then can consistent and coherent choices of priority be made between the distinct strategic alternatives available and the trade-offs they may require. One cannot plan for success unless one has defined what success looks like.