email icon Email this citation

Liberalism and Security: The contradictions of the liberal Leviathan 

Barry Buzan  and  Ole Wæver

Copenhagen Peace Research Institute
April 1998

1. Introduction

"Feare and Liberty are consistent"
 Hobbes, Leviathan, ch. 21

It is now commonplace to observe that the agenda of international security has expanded. Security policy, and thus security studies, are no longer only, or even predominantly, about military and "high politics" issues, but also about economic, societal and environmental ones. Some people in the field regret this turn, and others embrace it, but few bother to ask why it happened. To the extent that this question is asked at all, the answer is posed in simple terms either as an accumulation of random events (economic globalisation, the rise of environmental concerns), or as an effect of the ending of the Cold War (the relative decline of military priorities leaving more room at the top for other issues to claim security status). We think that there is an additional, deeper, story to tell, and one that has important implications both for how security policy is made and for how analysts understand and approach the new security agenda. This story is about the relationship between liberalism as a practicing political ideology, and the agenda of international security. It is a dialectical story in two parts: "classical liberalism" (sections 3 and 4) and "real existing liberalism" (section 5).

The first cycle of the dialectic covers the rise and eventual success of classical liberalism. It starts by being an ideology of opposition to the dominance of monarchism and mercantilism. Then it gains increasing strength and influence as the ideology of successive hegemonic powers (Britain and US), and becomes the main opponent of two totalitarian bids (fascism and communism) to become the hegemonic ideology of industrial society. The Cold War period saw some surprising twists in the attitude of liberalism to security with first a broadening of security to include economic and ideological concerns (manifested in the launch of the concept of national security), and later during the high Cold War an intense militarization of politics, which in its own way went against the liberal tradition. It also saw a separation and professionalization of security and international political economy (IPE) considerations into distinct fields. By the end of the Cold War, the classical liberal project had not only seen off all of its rivals, but also successfully established an extensive zone of peace (in the form of a security community centered on the Western powers and Japan) within which the problematique of security defined as war had largely been solved.

As far back as Hobbes's Leviathan, classical liberalism was a project for "desecuritisation" - understood mainly as reducing the number of issues over which force could legitimately be used. Liberals wanted to restrict the rhetorics of threat and survival to the military sector and the state, in order to open up space for a civil society in which individuals did not deal with each other in a security mode. By the nineteenth century this project had expanded from the domestic to the international realm, with liberals aiming to eliminate war internationally. This project linked economics, security and the state into an integrated interpretation, using the still familiar arguments of free trade (ie. economic interdependence) and democratisation (i.e. democratic peace). Its aim was to maximise openness between states and societies, and to avoid a spill-over of security behaviour into sectors other than the military one. In the classical liberal view, security logic pointed towards the economic closure of mercantilism, which was precisely what liberals wanted to eliminate. In this part of the story the aims of liberalism eventually come into line with its effects. What starts as a liberal theory of international peace eventually delivers the goods for a large and expanding group of countries.

The second turn of the dialectic starts from the ending of the Cold War, when "real existing liberalism" is the hegemonic ideology, substantially in operation for most parts of the planet, and intellectually dominant in the deep Gramscian sense of being widely accepted as the truth. Here we focus on the powerful irony that successful liberalism, especially in its economic practice, appears to be the main driver of a strong movement to securitise a wide array of economic, societal, political and environmental issues, as well as traditional military ones. Here the emphasis is on a disjuncture between the aims of liberalism and its apparent effects. We argue that the very success of the classical liberal project of desecuritisation is the main cause of a resecuritisation that in some ways threatens the foundations of liberalism itself. In other words, the new security agenda is not just a random accumulation of issues somehow built up in response to a variety of different events, or a reaction to the end of the Cold War. Instead, it can be interpreted in good part as a systematic response to the pressures generated by the triumph of liberalism. It is liberal economic practice that both integrates and impells the new security agenda. That line of thinking then leads to a strong call for reintegrating security studies and IPE, not just because their agendas overlap here and there, but because there is now a deep and systematic connection between their fundamental subject matters.

2. Liberalism and Security

Before exploring this story, we need first to look at its two central ideas: liberalism and security. These two concepts are commonly assumed to belong to separate research agendas, security mostly to realism and strategic studies and liberalism mostly to IPE. They are even used to define contending approaches to the study of International Relations (IR). Since the onset of the Cold War, liberal theory has downplayed security and security studies mostly ignored or dismissed liberalism. But although the relationship is turbulent, liberalism and security are in many ways mutually constitutive, both in theory, and in the ever-changing interplay between "real existing liberalism" and actual security practice. Their strained relationship has deep roots. The realm of security threatens liberalism, since the motives that propel actors in their security behavior are what liberal theory finds most difficult to explain, and what liberal practitioners see as most dangerous to their projects of liberty, openness and economic progress. Equally, liberalism challenges the logic of security by asserting that the supposedly permanent realist world of fear and "defensive positionalists" can in fact be not only alleviated, but possibly even replaced altogether. If states act according to a liberal logic of maximising absolute gains and generally prioritise economics over politics, then the war problematique of security will become marginal, and both security studies and the security institutions of the state will eventually become redundant.

Despite its common usage, liberalism is a surprisingly slippery concept. We understandit as a package of ideas and doctrines that are often mutually supporting, but sometimes contradictory. 1 Many elements of this package have deep roots, but it only came together as a unified political program in the 18th and 19th centuries. Its core idea is individualism, from which most other aspects of it derive. Individualism leads to private property, markets, the rule of law (within and eventually between states), and a preference for maximising the role of civil society while minimising that of the state. Individualism also leads to rationalism, secularism, tolerance, belief in progress, and belief that these ideas are universally good and universally applicable to the human condition. Given these values, liberalism is opposed both to autocracy (its original opponent), and to collectivist, totalitarian models of modernity that challenge its self-perceived right to inherit the world (its challengers during the twentieth century). As a rule, liberalism is hostile to the balance of power, war and militarism, preferring more rational, legal and institutional approaches to international order.

But although there is a certain coherence to this package, its internal contradictions mean that in practice liberalism is mostly talked about in terms of different strands. "Economic" or "pure" or "neo" liberals tend to put markets first, assuming that democracy, peace and individualism will eventually follow from the operation of a market economy. 'social" or "compensatory" liberals tend to put human rights first, often accepting restrictions on the operation of the market in order to meet basic human needs in a more direct and immediate way. This split means that liberals can end up in different places on the spectrum of party politics. Economic liberals can make alliances with some conservative and even some authoritarian elements, though liberalism is in contradiction both with the paternalistic and fascist strands on the right. Social liberals can make alliances with social democrats, though liberalism is in contradiction to the collectivist principles of both socialism and communism.

Our focus will mostly be on economic liberalism, partly because it is that strand which has become ideologically dominant, but mostly because economic liberalism operates much more strongly at the international level, which is where the international security agenda is largely located. Social liberalism is stronger in the domestic realm, though its human rights agenda does spill over into international relations. Economic liberalism is also strong in the domestic realm, but its market logic has also become very powerful in world politics. Our concern is with liberalism as a practicing political ideology, and especially as a hegemonic ideology, and we are primarily interested in the international rather than in the domestic application of that ideology. It is not our intention here to engage with the extensive philosophical debates about liberalism, though it is difficult to introduce the formation of liberalism without dealing with Hobbes, Locke, Mill and such figures. We use academic writings mainly as illustrations of the different twists of the changing relationship between liberalism and security. We study liberalism less as philosophy or theory and more as actually implemented political and economic practice.

Section 3 of our story is about British liberalism, which is the dominant locus up to the First World War. In this period, the liberal project is desecuritisation, and the security problematique is defined in terms of war. Section 4 is about American liberalism which is dominant after the Second World War. In this period, the project is securitisation (and the invention of "national security" as a concept) in response to the threat posed by the Soviet Union and communism to liberal hegemony. The messy transition period between them during the interwar years is covered only briefly. Section 5 is about global liberalism, which still has a strongly Anglo-Saxon flavour.

As if the difficulty in pinning down liberalism were not already great enough, there is the additional problem (especially for North American readers) created by the idiosyncratic development of terminology in the US. In today's American party-political debates "liberal" has come to mean almost the opposite of what it has traditionally stood for, or at least to focus almost exclusively on the social liberal agenda. To be called a liberal in American political discourse at the moment means to be anti-market, pro-big government, and in some ways (such as support for positive discrimination programs) anti-individualist. 2 It means what in Europe would be covered by 'social democrat" or the milder forms of socialism. This curious semantic turn is not only very atypical of the way the term has traditionally been used, it is also particularly misleading in relation to the generally liberal character which is widely acknowledged as the most distinctive feature of US politics and society. In this more general sense, liberalism in the US is not so much a position within a wider political spectrum, as it is the overal political consensus, the common platform on which most people stand even when they disagree. From an outside perspective, the striking thing about American politics is how much of it takes place within a liberal framework. The great majority of American politicians and political actors are "liberal" in the classical sense of individualism, market economy and their basic concepts of politics, society and economy.

Although much criticized, this argument was forcefully made by Louis Hartz in his 1955 book The Liberal Tradition in America. Liberalism became instantly hegemonic in the American polity, because (as noted by Tocqueville) the US was born without a true aristocracy, and its fight against the English King was carried out on the basis of those English principles and doctrines most supportive of the American case: the radical liberal ones. America was thus born liberal according to Hartz, and a number of developments reinforced this character. The expanding frontier, and the constant possibility of people moving to a place where they could have their own chance, both reinforced individualism. Neither trade unionism nor socialism ever developed as strongly in the US as they did in most other leading Western democracies. Despite this, the term "liberal" only made it strongly into American politics during the 20th century, most notably when President Wilson shifted his self-labelling from "progressive" to "liberal" between 1916 and 1917. At stake was Wilson's move towards entering the first World War, which would have been hard to reconcile with the progressive label given the anti-war sentiments prevaling among progressives. Ironically, the liberal label was introduced into American politics to focus attention on international affairs, and was therefore attacked by Randolph Bourne as "war-liberalism" (Green 1987: 76-85). Probably due first to the usage of the label for regulation-oriented progressives and later the absence of a powerful non-communist left-wing, liberalism gradually became the standard label for "left wing" leading to the current conceptual muddle.

In this paper it is the classical sense of liberalism, and the mainstream liberal character of US politics and society, that interests us, and not the misleading term that has evolved from contemporary American domestic political usage.

Security we understand broadly as being that special type of politics in which specified developments are constructed as being existential threats to core values and/or assets of human collectivities. Security in this sense is not a synonym for "military". Security politics can develop in relation to economic, political, societal and environmental issues as well. To construct something as a security issue means to call for exceptional measures to deal with it. Securitisation is therefore not just a call for political priority, but, if need be, for permission to break the normal rules of politics (i.e. by using force, by taking executive powers, by imposing secrecy). In contrast to the liberalism's desire to open up human life to an ever wider range of contacts and interactions, security is thus mostly about calls for closure against things perceived as threatening. When security is understood in this way, liberalism can be seen as a general project for desecuritisation. It seeks to narrow the range of things seen as threats, and to enlarge the realm of "normal politics" both within and between states. 3

In this constructivist view, the definition of what is legitimately part of the security agenda and what is not, varies over time as well as between different societies. The dominant threat and the responses to it may well be military, but they may also be in other areas. Migration, crime, pollution, flooding and competing cultures and identities can all be constructed as security issues (ie. securitised) that may be as, or more, important than military ones. Conversely, as witnessed most recently with the ending of the Cold War, even very intense military securitisations can quite quickly undergo comprehensive desecuritisation. Thousands of nuclear weapons still exist in Russia and the US, but for the most part these are now treated as "normal politics" between the two countries.

What we want to examine is how the relationship between liberalism and security has evolved as liberalism has risen from being a radical doctrine of opposition at the beginning of its political life, to being something close to a globally hegemonic ideology at the end of the 20th century. In particular, we are interested in the interplay between liberalism on the one hand, and the processes of securitisation and desecuritisation on the other. How in practice has liberalism shaped what is and what is not seen as a legitimate security issue? Given the nature of liberal ideology, one might expect that the practice of liberal politics would work to narrow the scope of security, and to constrain the power of the state to invoke security. In fact, the picture is rather more complicated. One possibility is that the general aims of liberal ideology and the effects of liberal practice line up in such a way as to promote the expected desecuritisation. It can be argued that they have done so within the expanding sphere of the West during the last half-century. But liberalism, like all ideologies, is vulnerable to challenges, and the Cold War provides a good illustration of another possibility - the quintessentially liberal state, the US, leading a thoroughgoing and broadly successful campaign for securitisation of the threat from communism. A third possibility is that while liberal aims push towards desecuritisation, the effects of liberal political and economic practice in fact generate pressures that enhance processes of securitisation. This is perhaps one way to read the experience of the interwar years, and we argue that in a radically different form it is what is happening now.

From the above arguments it follows that liberalism can have two approaches to securitisation and desecuritisation. The first, and simplest, is illustrated by the ending of the Cold War. Liberals no longer feel that their project is threatened, so they wind down the process of securitisation that they themselves created. In principle, liberals can move back and forward along this spectrum according to how they perceive the prevailing historical circumstances (threatening or not to the liberal project). The second derives from liberal ideology, involves two simultaneous moves, and in principle should only go in the direction of desecuritisation. The first move is to take economic, political and societal relations out of the realm of security, leaving only military issues as a legitimate subject for securitisation (thus Adam Smith's much cited remark about defence being more important than opulence". The second move is to use the increasing openness in other areas (free trade, democratisation) to progressively eliminate the sources of military insecurity. If successful, this scheme should eventually lead to security per se becoming a residual, peripheral part of political relations among states and peoples. The force of this second line is illustrated by its success in redefining political relations amongst the Western states. But this too can go wrong. As noted above,we will argue that the effects of liberal economic and political practice can in fact work to reopen processes of securitisation even in the non-military areas previously thought most thoroughly desecuritised This raises some serious questions, not just for security analysts and policy-makers, but also for the intellectual and political custodians of liberal ideology.

3. Containing Security: Classical Liberalism as a Strategy for Desecuritisation

Liberalism from Hobbes onward can be interpreted as a project for desecuritising ever-wider realms of human relations. In the liberal vision, economic matters are to be dealt with on purely economic premises, ideas exchanged freely without intervention of force, and state intervention reserved for ultimate threats such as foreign military attack or domestic revolt. Force and dramatic "necessity" are removed from the normal relations among citizens, and thus they are left to deal with each other without drawing on securitisation - or drawing their swords. In a sense, Leviathan is a proposal for desecuritising civil society, creating islands of civility within states, but not solving the problem of war between them.

The liberal aims of liberty, individual security and economic growth are furthered through a sequence of desecuritising initiatives which in turn produce an increasingly clear and concentrated definition of what security is, where it is located institutionally and who is legitimized to act in its name. This process can be seen as four phases, from Leviathan, through the 17th to 19th century defence of liberty against the state by rule of law (Locke, Montesquieu, Mill, Kant and Bentham), and the laissez-faire logic of Adam Smith and Manchester liberalism, to the explicit 19th and early 20th century attempts by liberalism to subdue security dynamics internationally.

In an IR context, it might sound puzzling to list Hobbes as liberal after the insistent claims made for his lineage by realists. But the early history of liberalism is most easily understood as elaborations and revisions on Hobbes. 4 His premises were individualistic (not organicist or romantic like the German ancestors of realism). Hobbes's starting point is that the individual has a right to self-defence, but that the individual pursuit of self-preservation is vulnerable. Therefore, it became rational to hand over to a Sovereign both the calculation and implementation of what would be best for matters of self-preservation.

This submission of will and judgement to Leviathan covers only matters concerning "the Common Peace and Safetie", whereas individuals would still be free to make their own judgements in all other matters. Of course in practice, the sphere of liberty for the subjects was decided by the rules of the Sovereign. 5 An enlightened Sovereign keeps large the sphere where subjects are free to act and limits himself to actions necessary to defend himself and to do what was necessary for his subjects" preservation. 6 By concentrating security in the state, Leviathan protects the subjects and thereby removes their incentives and ability to use force amongst themselves. 7 This was seen as the crucial gain, because the violent religious and political controversies of the time threatened to disrupt all spheres of life. 8

Starting from individual security, while ultimately vesting the authority over security in the sovereign, was the common premise for early liberal thinkers from Locke through Leibniz and Montesquieu to Condorcet. 9 "Individual security, in the liberal thought of the Enlightenment, is thus both an individual and a collective good. It is a condition, and an objective, of individuals. But it is one that can only be achieved in some sort of collective enterprise". 10 John Locke's Two Treatises of Government from 1689 is the work most often seen as installing liberalism. Since then, liberals have had an ambiguous relationship to Leviathan. While more or less explicitly accepting the basic Hobbesian argument that a social contract constructing a commonwealth was necessary or at least preferable for security and thereby liberty 11 , they also found it necessary to tame and constrain the state. To enable individuals to exercise their personal freedom, a state was necessary to defend both towards the outside and to defend individuals against each other - but too often Leviathan itself became the threat to individual liberty or even life. This dilemma defined much of liberal political theory throughout the 18th and 19th centuries. With Locke, Montesquieu, Kant, Bentham and J. S. Mill the state is constrained by the rule of law, constitutional government and division of powers.

With this taming of the state, a stronger, more specific formulation of the exceptional case had to be found. From the mid 19th century, the liberal state enters a juridical self-limitation, but one that "is balanced by the designation of a range of "governmental acts" which are immune to legal challenge. This juridical reserve area of executive power is (...) the qualification which (...) calculations of security impose as a condition for the political feasibility of a liberal democracy." 12 A more precise legal concept of "emergency conditions" replaces the vaguer, more general right for the state to break the rules simply by an appeal to necessity (and raison d"etat), 13 and this move constitutes a large step towards the modern concept of security. Only when something is particularly radical - existential threats - is the state allowed to go beyond the ordinary rules. (The intriguing problem remains of how to deterimine when this line has been crossed. If the state holds the right to invoke security, then there is no check on abuse. Yet by definition it is almost impossible to construct the law so that it can anticipate and regulate the extra-legal episodes. 14

A third round of self-limitation by the state was driven by the new economic logic of laissez faire which gained ascendancy in the 1820s-30s. Adam Smith and other members of the Scottish Enlightenment made a case (in the 1770s and 80s) for the positive economic effects of a non-interventionist state. Whereas this is now generally seen as a narrowly economic theory, it was part of a general redefinition of the role of the state, an orientation towards governing through indirect means and securing the necessary framework for an unfolding of individualism. 15 The state should not simply maximize its benevolent regulation 16 but ensure an over-arching rationality (which it could do better if it did not directly try to organize and meddle). The nature of economic processes prevents the state from fully knowing and thereby steering them; a better result is achieved when rationality is delegated to the level of individuals pursuing their own gains (which the invisible hand channels to a public good). Central is the idea of a separation of economic and political spheres which enabled a self-regulated economy. 17 Laissez-faire was not natural. It was enforced by the state.

In this phase liberalism became closely linked to the state as the mechanism by which individual liberty and good government could be assured (Latham, 1996: 81-92). Security and liberty became closely linked. In contrast to the second phase, liberty is not just against the state, but a central element of how to govern. 18 Bentham went as far as equating liberty and security; "what many people thought of as liberty, he preferred to treat as security" 19 By this time, it is not appropriate to define liberals as "anti-state". The relationship of liberalism to the state is characterized by "the need to maintain a suspicious vigilance over government so as to check its permanent tendency to exceed its brief in relation to what determines both its necessity and limits - society". 20 Or in Judith Shklar's formulation: "No liberal ever forgets that governments are coercive" 21 .

These domestic developments of classical liberalism were increasingly accompanied by an explicitly international liberal security strategy. Liberalism can be seen as addressing the problem of war and international relations for two reasons. One is offensive or messianistic, to extend the logic of liberalism in correspondence with its universalistic pretensions. 22 The other is defensive seeing it as necessary for the survival of liberalism domestically, because war and militarism are the most efficient excuses for tyrants to encroach on liberty by referring to external threats, establishing standing armies, and mobilising the power of the state to stand above both individuals and society. 23 In both cases, the aim parallels the domestic one of removing violence and violation of individual liberty and to open up space for the operation of markets.

The free traders of the 19th century tried to extend the liberal scheme, but with a different logic from the one liberals had employed domestically. Then (as now) they wanted to desecuritise, and eventually demilitarize, relations between states not by constructing a global Leviathan, but by pursuing the logic of economic interdependence between states and democratisation within them. The basic liberal idea of openness in economics, politics and society means that states and societies agree to narrow the range of things that they treat as security issues. This is in contrast to closed mercantilist and totalitarian states that securitise everything from pop music and ship-building to boots and bibles, in the process both impoverishing themselves and threatening others, as illustrated most recently by the former Soviet Union. The idealistic goal of the liberal project is first to squeeze security into the military sector, and then (hopefully) to erode the legitimacy of the use of force, eventually eliminating security issues. Free trade was the most consistent and forceful suggestion of the liberals, but liberal thinking on international relations emphasized other instruments in varying mixes: public opinion, reason, internationalism, interdependence, disarmament, information, democracy, and international organisations. 24

The liberal anti-state orientation is clearer here than in domestic politics. Internationally, the state is the problem: international affairs will be better the less involvement there is by princes, diplomats and generals. Liberal international relations is about complete de-securitisation, not only, as in the domestic sphere, the concentration of security in an ever more precise instrument. In Cobden's famous words: "As little intercourse betwixt the Governments, as much connection as possible between the nations of the world" 25 . Liberal anti-statism in the international realm sits awkwardly with its increasing parallel commitment to the legitimate role of the state at home, a contradiction that reappears strongly towards the end of the twentieth century.

was during this phase that the still familiar liberal package of democratization, free trade, international law and international organisation got established. 26 It had some trial runs during the 19th century even though most international practices remained realist. The first great experiment with international organisation did not take place until after the trauma of the First World War had made war itself into the main threat to liberal civilisation. The story of the League, its over ambitious commitment to collective security, and the misplaced faith of liberals in the power of public opinion is well known. 27 So also is its calamitous failure, which opened the way for forty years of realist dominance in IR. This history can be read as another instance in which liberal aims (collective security, peace, free trade) fell out of line with the effects of liberal policy (financial collapse, neomercantilism, spiralling security dilemma). Less well appreciated is the liberal IR literature of the interwar period, which actually had rather sophisticated theories of interdependence in which liberalism and modernity as historical processes were theorised and their potentials for new security strategies investigated. 28 Also less well appreciated were some of the contradictions within liberal policy at this time, such as how the strong demand for control of the arms trade squared with the general commitment to free trade and minimal state interference. These contradictions re-emerged as part of the security problematique caused by liberal praxis in the late twentieth century.

Through all four of these phases, classical liberalism reveals itself as a strategy for desecuritisation. Ever larger areas of social life are to be freed from the drama and logic of necessity characteristic of security. The concept of security is thus kept narrow: only that which is about military matters or other physical violence justifies extraordinary intervention. At the same time, the liberal way of thinking about international security (and also the realist/mercantilist and Marxist ones) reveal that there was no question that the subject matter of international political economy and international security were closely intertwined. From Cobden, through Hobson, Lenin, and Polanyi, to E.H. Carr, the nature of the state and the organization of the international economy are seen both as central determinants of the problem of war and as areas in which solutions to it might be found. Given this antecedent, how did security studies and IPE come to be seen as virtually separate realms during the Cold War?

4. The Cold War and the Liberal Security State

In the period following the Second World War, the torch of liberalism passes firmly from British to American hands. It is American liberalism that takes charge of the Western side of the Cold War, and here we see liberalism in robust securitising mode, generating a crusade in defence both of its global achievements to date, and its right to dominate the future of humankind.

Phase 1 - the 1940s: security launched as an integrative concept

Because of the military fixation of security in the 1950s, it is often forgotten that the opening phase of the Cold War pointed in the direction of a more expansive idea. The new American concept of "national security" which crystalised during the 1940s sustained a close policy linkage between 'security studies" and "IPE" (though neither of these had yet come into being as explicit sub-disciplines). The National Security Act of 1947 integrated a wide range of activities from the perspective of security, and the National Security Council was set up "to advise the President with respect to the integration of domestic, foreign and military policies relating to the national security". 29

For this period, the relevant liberalism is that of the United States which became the main driver of security policies and thinking. Historically, American liberalism had had a problematic relationship with foreign affairs. Both geography and sentiment had allowed it to indulge in isolationism and a suspicion of standing armies, though it had not made it immune to the temptations of imperialism. 30 The American liberal tradition did not follow on directly from mainstream British liberalism. Instead it raised to international prominence a radical and partly conspiracy-theory based 17th -18th century liberalism that never gained hold on the British: more Milton than Locke. 31

This liberalism, hostile to sustained military mobilization but also aggressively opposed to authoritarianism and inclined to self-righteous single-mindedness, came to shape the form security took. More specifically, security was shaped by the contradictory pressures of reacting to Soviet Communism as a broad-spectrum external threat on the one hand, and containing the risk of domestic military threats to the liberty of American civil society on the other. The mid-1940s was the moment of ascendancy for the concept of security. The word security as such is old, 32 but in the US in the 1940s "national security" became what Yergin calls the "commanding idea", 33 the standard label for a whole field, previously discussed as war, foreign policy, defence, or military policy. It arrived with such force, that it 'seemed always to have been with us". 34 As Rosenberg points out, "national security" combines the best of two discredited terms: the too realist "national interest" and the too idealist "collective security": "To grasp the discursive importance of the new term "national security" in shaping how the "reality" of the Cold War came to be inscribed in both language and institutions, one need only imagine the fate of a "National Interest Act" or, even worse, a "Collective Security Agency"". 35 The writers who launched this new concept- Edward Mead Earle and Walter Lippmann in particular 36 - were interested in fostering a closer integration between military affairs and foreign policy. The US which had been used to relying on geography for defence, and been highly suspicious of peace-time mobilization in the name of war 37 had to be taught a more general involvement and preparedness through a new slogan: national security. As with realism, it became more dogmatic and programmatic in the US, where it was controversial, than it had been in Europe, where such methods were so traditional as to be almost commonsensical.

One of the pressures leading to the concentration on security as an integrative concept was the US's experience of coordination problems during the Second World War. 38 As a sense emerged that the post-war period would be a possibly lengthy struggle with a large Eurasian power, it was deemed necessary to design an institutional and conceptual framework for a coordinated effort both within the military sphere and not least between military, economic, foreign policy and domestic "information" policy. Along with the advent of nuclear weapons, this reinforced the opening already made during the Second World War for civilians to move into security studies. The security problem for the West was how to respond to a challenge from the Soviet Union that was not just military, but also ideological, social and economic. Melvyn P. Leffler and other historians of the Cold War have shown how in the early years the American leadership expressly did not see the Soviet Union as a military threat. 39 The Cold War was about a broad spectrum rivalry between two mutually exclusive systems of political economy over the future of industrial society. According to Latham (1995), the intense securitisation of the Cold War that developed 1945-7 was also strongly tied to the project of building a liberal Atlantic order, which was already well underway before the Soviet Union had been constructed as a threat.

American liberals faced the dilemma that not only was the free world threatened by communism, defence against it could threaten liberty too. 40 Laswell's 1950 study on National Security and Individual Freedom is indicative of the problematique of how "to attain a high level of national security without at the same time making an unnecessary sacrifice of individual freedom". 41 Huntington in 1957 stated "The tension between the demands of military security and the values of American liberalism can, in the long run, be relieved only by the weakening of the security threat or the weakening of liberalism". 42 Paradoxically, the containment strategy was often justified with the help of the garrison state argument. In the words of Truman: "If Communism is allowed to absorb the free nations, then we would be isolated from our sources of supply and detached from our friends. Then we would have to take defense measures which might really bankrupt our economy, and change our way of life so that we couldn"t recognize it as American any longer. (...)It would require us to become a garrison state, and to impose upon ourselves a system of centralized regimentation unlike anything we have ever known" 43 A preemptive geopolitics, a wide security strategy, should avoid an intense military mobilization of the US itself that would threaten liberty. National security is a policy on behalf of liberty, but on the other hand, security policy threatens to undermine liberty. Liberty is both a motive for and a moderation of security activities.

In this first phase, the broad spectrum concept of security shows in the use of economics as a security instrument, not least in relation to Europe (from the settlement of the Second World War on a "no reparations" basis, through the Marshall plan and the Truman Doctrine, to support for European integration). 44 Another clear illustration of the effects of a wide concept of security was its use on domestic political dangers, most significantly in the form of McCarthyism. 45 But perhaps the strongest demonstration of wide spectrum liberal security logic was the US commitment to the idea of free trade supported by international economic regimes as the key to avoiding the conflictual political economy of the 1930s: Their belief was that: "if goods can"t cross borders, soldiers will". 46 Contrary to Polanyi's predictions, a new liberal, international order was established after the war, however, not the same as that of the 19th century. According to Ruggie, "embedded liberalism" was essentially a compromise: "unlike the economic nationalism of the thirties, it would be multilateral in character; unlike the liberalism of the gold standard and free trade, its multilateralism would be predicated upon domestic interventionism." 47 By shielding domestic stability, embedded liberalism rested on a wide concept of security, and the American-led international economic order was explicitly designed to achieve security goals.

Phase 2: the 1950s, 60s and 70s - the militarisation of security

The onset of the high Cold War saw a sharp disjunture with the early post-1945 construction of an inclusive concept of security that was sensitive to the contradiction between the demands of military defence on the one hand, and the threat posed by militarisation to the values it was designed to defend, on the other. Indeed, in comparison to classical liberalism, American crusading liberalism virtually abandoned the impulse for desecuritisation, subordinating it to what was constructed as an existential threat to the liberal project of the most extreme sort. During this phase, security was elevated to a leading position from where it coloured all other aspects of politics and society - exactly what liberalism traditionally wanted to avoid. The wide conceptualisation of security established in the US during the opening moves of the Cold War quickly narrowed down to an almost exclusively military focus. The tight linkage of military to security dominated the discourse right through the 1950s, 60s and 70s. It became part of the realist orthodoxy that security was about military issues, captured in the realist distinction between "high" politics (i.e. military-political strategic) and "low" politics (i.e. economic issues and most international organizations). 48 With the (re)emergence of thinking about interdependence and international political economy (IPE) in the late 1970s, this division became institutionalized. On the one side was an increasingly assertive IPE, whose practitioners were bent on using the leverage of current affairs to raise their agenda into the realm of "high" politics, and whose theories of complex interdependence (Keohane and Nye, 1977) created a subsystemic theory of liberal IR that marginalised the use of force. On the other was security studies, whose exponents remained largely locked into realist views of high politics, and a Cold War military agenda. For a time they existed as almost separate universes of study. Relations between them were at best indifferent, often dismissive, and at worst downright hostile (over competing claims to represent the real "high" politics).

There are two ways of explaining the depth and intensity with which the concept of security became militarized during the Cold War, one technological, the other ideological.

The technological explanation rests on the peculiar and dramatic impact of nuclear weapons and their delivery systems during the 1950s, 60s and 70s. 49 These weapons themselves transformed the prospect of war by raising the possibility that between nuclear armed states, victory would be indistinguishable from defeat. The rapid breaking of the US nuclear monopoly not only locked the two superpowers into an intense military competition, but did so at an early stage in the development of the technology itself. Consequently, throughout these three decades, radical changes in the performance variables of the main military systems meant that the normal tensions of military rivalry were greatly amplified by the pace of technological change. Strategic theories were forced to keep up with the ever-changing balance of capabilities created by new technologies and deployment decisions, and this made deterrence theory an understandably obsessive focus of concern.

Not until the late 1970s did nuclear weapon technology mature, allowing deterrence theory to mellow into both baroque complexity and existential simplicity. The possibility of defences against ballistic missiles briefly posed the threat of a second round of technological transformation, but the cost-ineffectiveness of this option pushed it onto the back burner. 50 Nevertheless, for thirty years the evolution of nuclear weapons and their interplay with US-Soviet rivalry provided ample justification for an extreme obsession with military issues - an obsession shared by both strategists and peace researchers. Other elements contributed to this militarization (the rise of the Soviet naval challenge, the perennial problems of NATO's central front, the success of guerrilla warfare in the third world), but the awesome destructive potential of nuclear weapons is the main technological key to understanding the depth and sharpness of the military turn in security studies during this phase.

The ideological part of the explanation rests, ironically, on the strength of American liberal reaction to the Soviet threat. As McKinlay and Little explain it, American liberals came to see totalitarianism in general, and (after the defeat of fascism) communism in particular, as an existential threat to the liberal project itself. 51 This was in line with earlier liberal crusading against autocracy (another closed, anti-individualist and usually anti-market political form). Totalitarianism threatened what liberals saw as the natural right of their superior form of political economy to inherit the earth. This zero-sum view mirrored the Soviet perspective of communism as the wave of the future, the interplay between the two quickly cementing in place a deep security dilemma. The Soviet Union was not just a military threat to the West. It was also a rival system of political economy whose basic principles were totally antithetical to both markets and individualism. Because of this profound incompatibility between liberalism and communism, the harder sort of American liberal always saw the objective of the Cold War as being the defeat of the Soviet Union and its ideology. Containment was not just a holding action, but a long-term strategy designed first to check Soviet expansion, and then progressively undermine the Soviet Union and confront it with superior Western power. This view underpinned the crusading element in US Cold War policy. It was not prepared to risk a premature war or the possibility of Soviet nuclear retaliation. But it pressed relentlessly for military superiority, warfighting options and escalation dominance, and resisted detente, arms control, the doctrine of mutually assured destruction (MAD), and anything that would allow the Soviet Union to be seen as equal, normal or legitimate. Reagan's rhetoric of "the evil empire" showed the continuing strength of this view more than three decades into the Cold War, and his ideas for grandiose defences against ballistic missiles were perfectly in accord with the hard liberal desire to escape from the paralysis of equal vulnerability imposed by mutual nuclear deterrence.

This liberal crusade remained a strong, though not always dominant, feature of American Cold War policy right into the late 1980s, when the arrival of Gorbachev provided it with the intellectual and political surrender of the Soviet Union. When combined with the nuclear obsession, the crusade against communism took liberals far away from their normal stance of deep skepticism about the balance of power, militarism and the state. Because of the existential nature of the communist threat to the liberal project, hard liberals were prepared to embrace, for as long as it took, the empowerment of the state and the pursuit of military confrontation. The wider political, economic and cultural elements of the conflict were always there in the background, but were overshadowed by the titanic military rivalry. During this phase hard liberals to some extent successfully stole the clothes of realism in presenting their case, in the process corrupting the understanding of what realism meant. Real realists such as Morgenthau, Kissinger and Waltz were prepared to deal with the Soviet Union as an ordinary state. They warned against the dangers of overstretch and self-weakening, even for so great a power as the US, that were inherent in an open-ended, crusading confrontation on a global scale. The softer sort of liberals were also conspicuous in promoting arms control as a way of reducing both the domestic and international dangers of intense military competition.

Mainstream liberalism was shaped by the Cold War, often in surprising ways. According to Arblaster, liberalism had already in the interwar years moved closer to conservativism. Many leading liberals (Orwell, Forster, Moore, Popper - and sometimes Russell) reacted to the First World War and the age of strong ideologies by defining liberalism by its doubt and tentativeness, its resistance to beliefs and fanaticism 52 - a thread which had been in liberalism from the start 53 , but previously combined with a rationalist belief in reform and progress. This definition of liberalism helped to single out communists, socialists, and radicals as beyond good liberalism 54 , paving the way for the "dedicated, quasi-religious anti-communism (...) Cold War liberalism" 55 Cold War liberalism shifted its terrain from philosophy to social science and thereby changed ethical ideals into political science definitions of liberalism and democracy corresponding to existing US realities. This transformation of liberalism had the significant impact on security that since the ideals of liberalism had been realized in the West, the liberal programme became one of protection, of defending against "totalitarianism" from without and radical "extremism" from within. The priority was no longer a vigilance against its own excesses, but a mobilization against foreign threats. 56

Taken together, these technological and ideological factors explain how it was that security became so strongly linked to the military sector as to make such a linkage seem natural, timeless, and beyond question - even though it was in fact quite shallowly and recently rooted, and quite at odds with the liberal thinking that preceded it. During this phase only one element of classical liberal security theory, functionalist integration strategies (cf. Deutsch, Haas), survived at the margins. But this theory was largely confined to the case of Western Europe, which was itself so deeply embedded in the strategic division of the Cold War as to make it an exception rather than a general model for IR or security studies. And because European integration became seen increasingly as a mainly economic affair (thus providing proof of its own desecuritising success), the security theory was largely pushed into the background. The separation of political economy from security that prevailed during this phase also had (for the West) the politically useful side-effect of legitimizing the post-1945 American/Western imperium, which operated on the demand for access rather than in the traditional European style of direct control. For many states and peoples in the periphery of the international system, the attempted liberal desecuritisation of the political economy was itself a security issue. Liberal states were able to delegitimize the non-military security claims of other actors, in the process subordinating them to the "normal" politics of the market economy and pluralist politics. By itself this situation justified a wider perspective on security, but only the voices of the weak calling for a new international economic order supported it, and it was largely drowned out by the superpower confrontation.

These historical developments took place during a crucial formative period of academic IR, when pre-war liberal idealism was banished, and realism established as the orthodoxy. The high Cold War, with its extreme focus on military security, thus helped to cement in place a polarization between security studies and IPE that in almost any other time during the last two centuries would have seemed bizarre, but which during this time of extremes seemed normal.

Phase 3 - the 1980s: the emergence of the new security agenda

By the 1980s, the dominance of military-political issues as the centre of security concerns was being questioned in several ways. The technology of nuclear weapons was largely mature, and deterrence theory had reached a point of intellectual and emotional exhaustion. 57 Except for a last surge of energy during the early 1980s caused by Reagan's 'strategic Defence Initiative" (SDI) and "Cold War II", the technological driver sustaining the militarization of security in the West was beginning to lose force. After the Vietnam war, there was an increasing tendency in the West to question whether war was a cost-effective method for achieving a wide-range of political and economic objectives. On top of this, the unfolding of Gorbachev's program during the later 1980s dealt a series of ever stronger blows against the ideological driver of militarized security, culminating in the dismantling of the communist challenge to market economics, and then in the dismantling of the Soviet Union itself. There was a growing awareness that war was disappearing, or in some cases had disappeared, as an option in relations amongst a substantial group of states. The core group of this emergent security community was Western Europe, Japan and North America. Once Gorbachev assumed power and embarked on an explicit desecuritising of East-West relations it became possible to think that the Soviet Union might also join this war-free sphere. During this process there were substantial moves towards arms reduction. If war itself was fading away as a possibility amongst many of the leading powers in the system, then realist assumptions about the inherent primacy of military security became questionable, and hard liberal ones about the temporary necessity for military containment became redundant.

Alongside the declining salience of the military agenda was the increasing securitisation of two issues that had traditionally been thought of as low politics: the international economy and the environment (more on both of these in section 5 below). In the case of the environment, the growing impact of humankind was transforming the natural environment from being a background constant into a foreground variable. Starting from 1960s concerns about pesticides, this grew steadily into a wide range of interconnected issues including climate change, biodiversity, resource depletion, pollution, and the threat from meteorites. 58 In the case of the economy, the securitisation process arose in part from the relative economic decline of the United States, and in part from reactions to the increasing liberalisation of the world economy, first in trade, and from the 1970s also in finance. In general, national economies became progressively more exposed to competition from other producers in a global market, and to ever more powerful transnational corporations and financial markets. This development gave rise to specific concerns about the ability of states to maintain independent capability for military production (and therefore mobilization), and about the possibility of economic dependencies within the global market (particularly oil) being exploited for political ends. There were fears that the global market would generate more losers than winners, and that it would heighten existing inequalities both within and between states (manifested at the top of the range by US fears of decline, and at the bottom by developing country fears of exploitation, debt crises and marginalisation).

During the turbulence surrounding the ending of the Cold War, the changing nature of the security agenda itself became a focus of controversy. The virtual collapse of Cold War military concerns by the late 1980s, and the proliferating attachment of 'security" to an ever wider range of issues, raised protests from the security studies establishment that the concept of security was becoming debased. 59 Traditionalists fought back both by reasserting conventional arguments about the enduring primacy of military security, 60 and by raising the charge that widening the meaning of security beyond the military sector invited intellectual incoherence. The key strategy was to allow widening only inasmuch as this could be linked to concerns about the threat or use of force between political actors. In a landmark statement of the traditionalist position, Walt argues that security studies is about the phenomenon of war, and that it can be defined as "the study of the threat, use, and control of military force". Against those who want to widen the agenda outside this strictly military domain, he argues that this:"runs the risk of expanding 'security Studies" excessively; by this logic, issues such as pollution, disease, child abuse, or economic recessions could all be viewed as threats to 'security". Defining the field in this way would destroy its intellectual coherence and make it more difficult to devise solutions to any of these important problems". 61 Walt 62 does allow "economics and security" into his picture, but only as it relates to military issues, and not as economic security per se.

As its main defence against the wideners, the mainstream security establishment thus focused on the charge of intellectual incoherence, and retreated into a dogmatic military definition of security. It is curious that they relied on this relatively superficial argument when a much more serious and powerful line is available. 63 Widening threatens the whole liberal project by bringing back into the realm of security many issues that liberals have sought, with considerable success, to desecuritise. To the extent that liberalism is defined as a desecuritising project, the unrestrained widening of the security agenda threatens both its conceptual foundations and its accomplishments. It is more than a little surprising that such a line of attack has not been used against the wideners, except in a limited way by Deudney 64 (and in our own previous reflections 65 ). The wider agenda certainly seems to be more vulnerable to excesses of securitisation than the traditional military one (which is vulnerable enough by itself if taken to extreme). In the immediate wake of the Cold War, the lesson from the Soviet Union about the massive drawbacks of excessively wide securitisation stand exposed for all to see. In this perspective, widening the security agenda can be cast as a seriously retrograde move. It threatens the hard won desecuritising achievements of liberalism, and perhaps even of the Hobbesian Leviathan, over the past three centuries, and is out of line with the strong liberal imperatives towards more openness in the post-Cold War world.

Seen not as a product of the Cold War, but as part of the liberal program of desecuritisation, the retreat of traditional security studies into the military sector makes clear sense. The liberal project to limit the scope of securitisation argues in favour of the traditionalists, with their narrow agenda, and against the wideners. Reserving security for the military sector has a pleasing "last resort" ring about it, and fits comfortably with the broadly liberal ideology that is now enjoying ascendance. Demilitarization by sector has been the characteristic liberal approach to desecuritisation, and in that sense traditional security studies can be seen, surprisingly, as one of its products (and not just of realism, as is generally assumed). For what is the traditionalist style of security studies about if not the isolation of the military sector as embodying 'security"?

There is a deep contradiction in this situation. On the one hand, the core security ideas of classical liberalism about the benign effect of free trade and democratisation stand triumphant. Now freed from the securitising imperatives of the Cold War, they dominate many of the most powerful states and societies on the planet. Although far from universally successful, liberalism has seen the rolling back of force in the domestic politics and economies of many states, the spread of democratic norms and practices, and the widespread adoption of more open economic relations between states. Within the Western core, there is a degree of international institutionalization and socialization that has virtually ruled out war amongst the states. By almost any measure, the liberal desecuritisation project had by the late twentieth century achieved spectacular success. The ending of the Cold War, and the surrender of communism to the market, underlined this success as much as it contributed to it. And yet on the other hand, this moment of triumph is accompanied by a durable and impressive movement to widen the security agenda across practically the whole range of human activity, seemingly bringing into question the very foundations of the classical liberal project. How is this apparent contradiction to be explained? Ironically, as we hope to show, it is the very success of the liberal project that now gives rise to the demand for a wider security agenda, for a reinvention of security in terms other than military.

5 "Real Existing Liberalism" and the New Insecurity

To understand liberalism as the cause of the wider security agenda, one needs to focus on economic liberalism, and especially on the praxis of the global market economy. Along with democratisation, market economics played a big role in liberalism's apparent success in solving the problem of war for a substantial part of the international system. But now that liberalism, and especially economic liberalism, has become both the hegemonic ideology and the dominant mode of organisation, a new framework for (in)security unfolds that is quite unlike the Cold War one. With liberalism defining many of the most important political and economic spaces on the planet, it simultaneously spreads the classical version of the liberal peace, and opens up a new set of insecurities. The new agenda of insecurity arises in large part from the operation of the global market economy itself. It is not an aim of liberalism (quite the contrary), but an often unintended and unanticipated effect of liberal policy in practice. It is partly about economic insecurity directly, and partly about the spillover of effects from the operation of global markets into the military, political, societal and environmental sectors. In addition, the rise of liberalism to hegemonic status increases the pressure that other liberal ideas, most notably individualism and human rights, put on societies that do not share them. The liberal peace is not universal, and in many respects it is imperial towards the remaining non-liberal societies. 66 Finally, democratization often adds to the widening of security because the new forms of insecurity will be felt and articulated by actors other than the traditional state representative. Because of the political and legal space opened up within and between states by liberal policies, these actors enjoy a higher degree of autonomy and freedom of action than before.

There is widespread agreement about the victory of liberalism - accompanied by disagreement about whether this is the "End of History" 67 or only a temporary win until new rivals arrive 68 This victory stimulated reflection on the international effects of liberalism not least through its linkage to the democratic peace thesis 69 . However, these discussions are often based more on liberal ideals than studies of "real existing liberalism," 70 and most of them therefore miss the linkage between liberal practice and the widening of the security agenda. One possibility, of course, is that the rather sudden evaporation of the Cold War, with its huge accompanying desecuritisation simply left more room for other issues to claim security status. This vacuum theory of securitisation suggests either that states have a general need to securitise something in order to justify their existence, or that the particularly heavy securitisation of the Cold War over four decades created a substantial inertia of security institutions and attitudes. This argument has some weight. The success of liberalism does create something of an existential crisis for the state: what is it for now that much of its economic role has been turned over to other actors, and (for some) its role of military defender has become largely redundant? And there can be no doubt that the Cold War did leave a heavy legacy of security institutions and attitudes that have trouble adapting quickly, or in some cases at all, to the global environment created by real existing liberalism. There is a security vacuum, and in part the rise of the wider agenda may be explained by this.

But we want to make a stronger argument than this. The wider agenda is not just a response to the victory of liberalism, it is in many respects a function of the way in which real existing liberalism operates. The argument is that there is a systematic and detailed linkage between how liberalism operates as a hegemonic ideology, and why there is sustained pressure to bring a wide array of non-military issues into the realm of security. The easiest way to get an idea of the insecurities that emerge under hegemonic liberalism is to survey them by sector, starting with the central economic issues, and noting how they spill over into all of the other sectors. 71

Economic insecurity

Actors in a market are supposed to feel insecure: if they don"t the market doesn"t produce its efficiencies. This fundamental axiom of liberal economic theory and praxis sharply limits the scope of economic security in liberal systems. 72 Liberals generally argue that the economy should be at the root of the social fabric, and that the market should be left to operate as freely as possible from interference by the state. The state is necessary to provide law and (politico-military) security, and to support the social fabric in areas where the market fails to do so. This means that under liberalism, firms should never be allowed to claim security status even if their survival is threatened. Individuals should only be allowed to invoke economic security if their survival (and not just their relative level of welfare) is threatened. Traditionally, the state was the main referent object, mobilizing whenever problems for specific sectors or firms were claimed to create unacceptable dependence or threaten the base for independent military power. Using arguments about national security served to legitimize protectionism or subsidies. Such policies have become difficult in a global economy defined by openness and de-regulation, and increasingly subject to WTO rules. Consequently the main referent objects of economic security under contemporary liberalism are the rules that create factor mobility within and among national economies. The economic liberal ideal is ultimately to dissolve national economies, with their exclusive currencies and restrictions on factor movement, into a global economy with relatively few restraints on the movement of goods, capital, services and (more hesitantly) people. The fundamental problem in this arrangement is how to maintain economic and political stability, and how to handle the widening gap between very rich and very poor, both domestically and internationally, that unrestricted markets tend to generate. This problem is made more difficult by the undoing of embedded liberalism and the removal of many of the welfare state mechanisms shielding society against the extreme effects of liberalization. 73

is these more system-level concerns that shape economic insecurity under liberalism. One set of issues centres on potential instabilities in the operation of global trade and finance, and the fear that these could generate a substantial collapse of the liberalising regime developed during and after the Cold War. The problem with trade is that the successful industrialisation of areas outside the West has not just increased the size of markets, but even more so has generated surplus capacity in many industries. It is hard to think of any mass-market product for which supply is not now in excess of demand. Surplus capacity intensifies competition, and results in deindustrialisation where older producers have become uncompetitive. 74 This globalising of economic efficiency is good for consumers, but it places tremendous pressures of adaptation on states and societies which have continuously to reconfigure the way in which they earn a living. The pain of continuous adaptation both for losers and for those trying to stay competitive can easily intensify pressures for protectionism, and these in turn threaten the global liberal trading order. The problem with finance arises from the liberalisation that has been underway since the 1970s, which has reduced the financial management tools available to states, and empowered global financial markets. This development exacerbates the trade problem in two ways: it weakens the ability of the welfare state to deal with the domestic consequences of intense competition, and it complicates trade and industrial policy by deranging exchange rates. More worryingly, the "Casino capitalism" 75 of highly leveraged futures and derivatives contracts involving enormous sums opens up the possibility of a major financial crisis resulting from the collapse of overextended credit arrangements.

These potential trade and financial instabilities are of course hypothetical possibilities latent in the way the liberal system operates, and there is a wide variety of views about whether their threat carries low or high levels of probability. If the international economy spins into a major crisis then it will be a central security issue. But if the measures taken to overcome or contain crises work, then most economic issues will remain off, or marginal to, the security agenda. Some writers are predicting a severe "time of troubles" ahead, 76 but then such views are a continuous presence on the landscape of punditry, and are not especially distinctive to this specific point in history. A global economic crisis may or may not come to pass, but the prospect of it, and awareness of its possibility, provides the basis for securitisation in the economic sector. The rise of economic security is therefore not just a throwback to classical mercantilist concerns about relative gains. It is a reaction against the various dangers inherent in the operation of a global liberal economy. 77

Perhaps rather less hypothetical is the problem of uneven development that arises not from the failure of liberal economies, but from their successful operation. The widening gap between very rich and very poor both within and between states opens the possibility that liberalism becomes seen less as about maintaining the rules that underpin economic efficiency, and more about protecting the position of the capitalist elite. As Richardson notes: "Liberal thought was, and largely remains, silent on the darker side of free trade, its integral association with power relationships." 78 During the Cold War, one aspect of this was the imperium of economic access demanded by the West, and the rejection of attempts by many in the Third World to pose this as a security issue in relation to their prospects for development. To the extent that liberal economics becomes identified as a mechanism of subordination or inequality, it becomes vulnerable to securitisation by those who are doing badly out of it. Avoiding that political turn requires liberal elites to avoid major collapses in the trade and financial regimes. It also requires a sustained commitment to economic growth in order to shrink, or buy-off, the circle of losers. That commitment creates a strong tie to the agenda of environmental security.

Environmental insecurity

In general terms, environmental security concerns the maintenance of the local and the planetary biosphere as the essential support system on which all other human enterprises depend. In terms of causality, some parts of the environmental security agenda are completely detached from the liberal economy, as for example the concern about space rocks smashing into the planet. Other parts are only connected indirectly, such as concerns about our position in the planet's natural cycle of ice ages, or about water shortages in parts of the Middle East and South Asia. But a substantial chunk of the environmental security agenda arises from the interplay between a liberal world economy and the sustainability of the planetary environment. In places this interplay is hard to disentangle from that between the growth of human population and pressures on the carrying capacity of the environment, but generally the picture is clear enough. On the one hand stand the ideology and praxis of liberal economics, with their emphasis on materialism and consumerism, and their political addiction to growth. On the other hand stand the finite capacities of the planet for food and energy production by land and by sea, and its fixed stock of non-renewable resources.

If the relationship between economic liberalism and environmental security was defined purely in terms of these two irreconcilable positions, analysis of it would be relatively simple. But the picture is complicated by two mediating factors for both of which liberalism can take much of the credit. First is the advance of technology, which allows growth to be intensive rather than extensive, and which can solve as well as create environmental problems. It is a commonplace to observe that new technologies are generally less resource and energy intensive than the ones they replace (think of computers over the last three decades), and that they are often less polluting (one way to reduce air pollution is to replace your old car with a new one) Second is the marked tendency of consumer capitalism, where it produces prosperity ad education, to slow down and even reverse the growth of human population. There is no advanced capitalist society in which excessive population growth is a problem, and some, most notably Japan, in which potential population shrinkage is becoming a worry. To the extent that economic liberalism supports democracy, it can also be argued that it opens the way for public pressure to be mobilised in favour of environmental concerns.

Nevertheless the central contradiction between economic liberalism and environmental security remains. At the centre of the part of the environmental security agenda driven by liberal praxis is the fear of either or both of a collapse of the achieved levels of civilization, and a wholesale disruption of the planet's biological legacy. The main issues are pollution, and human competition with other species for land and sea areas which they need as habitats, and which humans want for their own use or exploitation. Both are seen to be intensified by the materialism, consumerism and growthism of liberal economics (as they were also by the rapacious, but now largely extinct, inefficiencies of Soviet style command economics). The main threats that arise include:

As with some of the economic variables, what the environmental security priorities are will also depend on how a number of intrinsically unpredictable things work out. For example, some scientists argue, on the basis of drill cores from the Greenland ice cap, that serious climate change in the past sometimes occurred with great swiftness, major changes in temperature (and therefore in glaciation and sea level) occurring within a few years. If they are correct, and the global climate is more rather than less vulnerable to the industrial assaults upon it, then current observations such as the break-up of some Antarctic ice sheets could put environmental security at the top of the global agenda very soon. If they are wrong, environmental problems will remain localized, consisting of countries or regions with particular problems: sea flooding in a few very low lying countries; water sharing in the Middle East; nuclear accidents in Europe and suchlike. But either way, there remains a clear and substantial link between the ideology and operation of a liberal world economy, and the potential securitisation of a wide range of environmental issues.

Societal insecurity

Given the liberal concern to promote and liberate civil society, it is perhaps ironic that societal insecurity seems to be rising in a system of global liberalism. The praxis of economic liberalism again plays a major role, though in this sector so also do other elements in the liberal package, most notably the unresolved tension between the atomizing and homogenizing force of liberal individualism and the gemeinschaft forms of society that in practice came to constitute the political framework for most of political liberalism's aspirations. 79 In the societal sector, the referent object is large-scale collective identities that can function independently of the state, such as nations and religions. Given the conservative nature of "identity", it is always possible to paint challenges and changes as threats to identity, because "we will no longer be us". Whether an identity gets securitised depends on whether its holders take a relatively closed- or a relatively open-minded view about how their "we-ness" is constituted and maintained. 80

The two issues most commonly viewed as threats to societal security are migration and rival identities. With migration the fear is that identity will be changed by a shift in the composition of the population (e.g. Chinese migration into Tibet, Russian into Estonia). Rival identities can threaten in several ways. The penetration of values and practices from an outside culture can distort or even destroy an indigenous culture (fears of Westernization or Americanization in many places). Integration projects (the EU) or disintegration ones (Québec, Kurdistan) can threaten to submerge or fragment existing identities. Both types of threat have arisen from sources other than liberalism, but both are also now being promoted by it.

Liberalism is probably not the main cause of contemporary migration, but it is contributory, and the liberal quest to make borders ever more porous sets a background against which others can seek to securitise migration. The pattern of uneven development typical of liberalism (though not exclusive to it) encourages migration within and between states, as does the liberal view of the individual as being responsible for maximising his or her economic opportunities. Liberalism globalizes economic horizons, opens borders, and spreads information about conditions in other parts of the world. Economic deregulation can create demand pull in the richer areas for cheap, unorganized labour, and liberal political values such as human rights can create seemingly open-ended commitments to accept refugees from the overflowing parts of the world in which illiberal political and cultural practices are widespread.

Liberalism plays a more central role in generating societal insecurity through rival identities. Most obviously this happens when cultures feel threatened by Westernization or Americanization. In part this is simply the world-wide projection of culture from the powerful liberal core in the form of language, style, dress, dance, attitude, social values and suchlike. But in part it is also a spill-over from the praxis of the liberal world economy, which operates to undermine local cultures in several ways. It both encourages elements of cultural homogenization in order to create global brands and markets (pop music, clothes, drinks, films), and threatens to commodify distinctive elements of local cultures in order to sell them globally (most obviously in the context of mass tourism). The demand for "level playing fields" in trade negotiations very quickly runs into cultural issues that affect the traditional values and practices of society, as well as how its people are allowed to earn their living. In some of the more deeply rooted non-Western cultures (Islam, Iran, China), this package of individualism, materialism and commodification is viewed not just as "economics", but as a form of cultural assault. The challenge of adapting to the new rules of global competition uproots social groups, which then become open for articulation of societal insecurity in nationalist, chauvinistic movements. 81

Large problems of societal insecurity can also be found where the classical liberal project of desecuritisation has been most successful, such as in the integration project of the EU. When local military threats have dissolved into security communities, and national economies are bound together by a network of transnational and international rules, investments and organizations, old state structures that used to be justified for both defence and development become vulnerable to erosion from above. Regional integration is clearly part of the new rules of competition in a global, deregulating economy, and integration easily triggers societal insecurity. It can do so directly by generating a rival regional identity. But in practice (in Europe) the main effect is more indirect. The transfer of state functions to the regional level leaves the nation with less of its state shell. It thus becomes more vulnerable to either the invasive homogenizing forces of the regional market (as expressed in several national referenda, and Norway's continued rejection of membership), or the disintegrative forces of submerged identities seeking their own political expression (in the EU, most notably in Italy, Spain and Belgium; in North America, Quebec and Western Canada). Fears for the survival of the nation can lead to nations trying to restore the close link between nation and state and therefore opposing the integration project (Norway). Alternatively, a more systematic differentiation can be supported by both state and nation where cultural practices get the responsibility for defending national identity (Germany).

In an ideal liberal world, societal security concerns would be minor because people are tolerant, enlightened and open-minded. Tolerance and diversity should in principle prosper best under liberalism, but at the symbolic level, the opposite happens. As Pierre Hassner has pointed out, cultural homogenization might go hand in hand with a political search for differences (exactly to deny the actual similarity); this in contrast to the ideal many cosmopolitan intellectuals set up, where political culture is converging while cultural diversity is protected. 82 Much of societal security in the richest part of the world - from the overall orientation of the US after the Cold War to attitudes to the EU in the Northern part of Western Europe - is a conflict between a cosmopolitan, liberal, internationalized part of society and a more locally tied, communitarian resistance, i.e. more an opposition between universalizing and particularizing cultures than between different particularizing cultures.

Political insecurity

Political security is about relationships of rule and recognition, and concerns the organizational stability of states or other systems of governance, and the ideologies that give them legitimacy. With the waning of military threats, political insecurity comes more clearly into focus in its own right. Under a liberal system of international relations state sovereignty - the classical focus of national security - remains a central issue, but is threatened primarily from non-military directions. A variety of supranational referent objects also come into the picture. The EU, for example, can be existentially threatened by things that might undo its integration process. International economic regimes, and international society more broadly, can be existentially threatened by things that undermine the rules, norms and institutions that constitute them.

The diverse forms of political insecurity fomented by the ideology and praxis of liberalism rest on the general tension between liberalism and the state. As argued in section 3, there is a longstanding tension between liberal thinking on the domestic level, where the state is necessary to sustain both individualism and the market, and on the international level, where the state is mostly seen as being the problem in terms of war and protectionism. Economic liberalism is inherently antithetical to impermeable borders, and therefore to much of the territoriality that defines the state. Traditionally, liberals were prepared to concede the necessity of the state to deal with external (and internal) military threats. But as such threats have moved to the margins for many states, this principal justification for and role of the state have weakened. Pushed to extremes, as it might be under conditions of low military threat, neoliberalism can come close to being a doctrine of anarchism (albeit with property rights), in which social and economic order can be generated without the state. 83 This tension between liberalism and the state manifests itself in a variety of ways, depending on the type of state under consideration, and its place in the international system. In the space available we can do no more than outline the nature of this agenda using a simplified division of the contemporary universe of states into three types: post-modern, modern and pre- modern. 84

Modern states represent the classical Westphalian model, and until quite recently were the dominant type. They are defined by strong government control over society and restrictive attitudes towards openness. They see themselves as independent and self-reliant entities, having distinctive national cultures and development policies. Their borders mark real lines of closure against outside economic, political and cultural influences, and their sovereignty is sacrosanct.

Nearly all of the capitalist powers have now turned post-modern. They still retain the trappings of modernity such as borders, sovereignty and national identity, but don"t take them nearly so seriously as before. Post-modern states have a much more open and tolerant attitude towards cultural, economic and political interaction, and define a much narrower range of things as threats to national security. Post-modern states are pluralist and democratic, and civil society has as much or more influence than government.

At the opposite end of the spectrum, scattered throughout the third world, but most notably in Africa and Central Asia, is a set of states that can loosely be described as pre-modern, defined by low levels of socio-political cohesion and poorly developed structures of government. Some of these are pre-modern in the sense that they aspire to modernity, and are headed in that direction, but have yet to consolidate themselves sufficiently to qualify. Others are failed states, where the colonial transplant has broken down, and there is little other than external recognition to sustain the myth of statehood.

For post-modern states, political insecurity under liberalism is broadly defined by the hollowing out of the state. One driver of this erosion of sovereignty is the assertion of individual and group rights against the state (e.g. human rights).

Giving primacy to individuals not only undermines the legal claim of the state to sovereignty, but also provides strong foundations for challenging the right to nonintervention in domestic affairs, which is a cornerstone of the Westphalian international political order. If universal notions of human rights can be used to legitimate interventions, then the whole system of sovereignty as the basis of international political order is in question. In this perspective, the quite strong pressure building up within the more radical wing of security studies to demand that the primary referent object for security be individuals, not the state, can be understood as part of the liberal assault on the state. 85 The other main driver hollowing out the state is the globalization of the economy, which not only opens up state borders to movements of goods, capital, information and (to a lesser extent) people, but also empowers transnational firms and markets in relation to the state. Where interdependence has generated integration, most notably within the EU, the state already has to share sovereignty with other levels of governing institutions above and below the state.

One issue arising is the undermining of democracy. 86 Polanyi noted the contradiction between liberal economic praxis and democracy in the first attempt to construct a liberal world order. 87 Now power is being shifted away from the state and into economic institutions, such as markets and central banks, or political and legal ones, as in the case of the European Commission, the WTO and the European Court. Either way, pluralism is reinforced at the expense of democracy, and (more arguably) elite interests favoured over mass. Many of these institutions are less, or not at all, accountable to the people, for example the much talked about "democratic deficit" in the EU, and the fashion for moving control over exchange rates out of politically accountable finance ministries and into independent central banks. The technical case for this latter move is quite strong: low inflation is generally accepted as a good thing, and it is more likely to be achieved by independent banks than by government ministers subject to electoral and other pressures. But the political implications are disturbing. Precedents are being set for the idea that a global market economy (or a national one within it) can only work efficiently if major elements of policy are taken out of the political process. Democracy, in other words, is being chipped away by technocratic arguments that implicate the process of political representation as a main cause of economic mismanagement. This might be thought worrying regardless of the merits of the technocratic case. Among other things, it points to the exclusion from influence of those most negatively affected by liberal economic praxis: the unemployed and the economically downgraded in industrial societies, and the rural masses in developing democracies such as India. Since democracy has been a defining part of both the liberal project and post-modern states, this development raises fundamental questions. And since democracy more than free trade is the keystone to the liberal peace, 88 its erosion calls into question the foundations of one of classical liberalism's most astonishing and valued achievements.

A second issue is the impact of hollowing out on the provision of international order. Ironically, the same mechanisms that generate the liberal peace also seem to undermine both the capacity and the will of post-modern states (who comprise the bulk of the major powers) to take leading roles in the international system, making them "lite powers". 89 The resources of power are still present in postmodern states, but they are either not under government command, or not available for use because of public attitudes hostile to machtpolitik. By making the great powers "lite" the success of the classical liberal project threatens the traditional mainstay of international order.

The weakness of political leadership feeds into the question of the (in)stability of the liberal world economy. With the ending of the Cold War, longstanding concerns about the decline (or in some versions corruption) of the United States as a hegemonic leader 90 take on an ironic Leninist twist. The fashionable talk about competition for market shares 91 cuts just close enough to Lenin's ideas of imperialism, and the struggle to redivide a saturated global market, to generate a deep sense of unease. It also tends to undermine the main hope of political economy institutionalists, already understood as a fragile option, that shared understandings and ideologies, plus a collection of international regimes and institutions, might just be able to sustain the liberal international economic order on the basis of collective hegemony. 92 It may of course be the case that the "lite power" effect is largely confined to the military-political side of world order, and that the global economy will find ways of stabilising and institutionalising itself (such as the Bank for International Settlements) that do not depend on traditional great power hegemony. If this happens it would be a wholly new departure in the making and sustaining of world order, and one that would greatly strengthen the whole position of global liberalism.

For modern states, political insecurity under liberalism takes two forms: the threat of exclusion and the threat of inclusion. The threat of exclusion is most strongly felt by modernist states adjacent to the post-modern core, such as Mexico, Turkey, and many Central and Eastern European countries. For these states, exclusion means relegation to second class status, and denial of many benefits of membership in the core. The threat of inclusion is general to modernist states and arises from conflicts between their indigenous cultural and development projects on the one hand, and liberal values on the other. For many of these states, the price of economic and political relations with the West is exposure to demands for openness and 'standards of civilization" that amount not just to an assault on sovereignty, but in some cases (most notably Islamic ones) to an assault on identity. Recognition, aid and trade may be made conditional on legal reforms (particularly for property rights), human rights performance (reflecting liberal values of individual rights), currency reform, adherence to norms of multiparty democracy, and reduction of restrictions on the movement of goods and capital. As Iran, North Korea, Libya, the former Soviet Union, and to a lesser extent Argentina, China and India, can attest, the liberal core is actively hostile to rival modes of development. As Gilpin notes, liberals "cannot easily accept the doctrine of intellectual peaceful coexistence." 93

For pre-modern states, insecurity under liberalism is broadly defined by an inability to measure up to international standards of good governance. The threat is either that they will be demoted in the ranks of international society to some sort of trusteeship status, no longer recognized as legally equal and capable of self-government, or that they will simply be neglected and allowed to fall into chaos. To the extent that the strong and intrusive world system of a liberal order can be suspected of keeping periphery states weak, there is a particular link between liberalism and the plight of pre-modern states. 94 It may not be possible for some societies to develop modern (let alone post-modern) forms of social and political life when they are exposed to the rigours of a global market and the seductions of a wealthier, more powerful, global, culture.

Military insecurity

Military security is about relationships of forceful coercion, and concerns the two-level interplay of the armed offensive and defensive capabilities of actors in the international system, and actor's perceptions of each other's intentions. But with the stunning success of the classical liberal project, great powers no longer have to worry too much about wars against other great powers, and minor powers have much less worry in terms of imperial take-overs by great powers than has traditionally been the case. At best (or worst) what remained was a situation of "two worlds". 95 Rather than being a single politico-strategic space, with a single set of rules of the game, the international system has divided into a "zone of peace" and a "zone of conflict".

The zone of peace is defined by a post-modern security community of powerful advanced industrial democracies, and international relations within this world no longer operate according to old realist rules. Post-modern societies are in many ways not compatible with many of the traditional roles and behaviours of great powers. Individuals are no longer so prepared to die for their country. 96 Post-modern citizens trust their governments less, and are much more critical about state actions that might require heavy costs or sacrifices. The emphasis on individual and human rights means that foreign policy actions easily get bogged down in moral and legal debates. In the zone of peace, armed forces no longer face military threats from neighbours. The main de facto function of these armed forces is to support routine world order activities, such as peacekeeping or humanitarian intervention, that cannot be viewed as security in the sense of concerning existential threats to their states.

The zone of conflict comprises a mixture of modern and pre-modern states. In relations amongst (and within) these states classical realist rules still obtain, and war is a useable and used instrument of policy. In this zone, states expect and prepare for the possibility of serious tension with their neighbours. Some restraint is provided by deterrence (in a few places nuclear). But economic interdependence between neighbours is generally low, and the force of the "democratic peace" argument weak: populations can often easily be mobilised for war.

The two zones idea suggests that liberalism has largely eliminated the military security agenda in one area, while leaving it largely untouched in the other. But this is too simple a conclusion. Questions remain both about the impact of liberal political economy on "relationships of forceful coercion", and about how the two zones relate to each other.

The effect of liberal praxis in hollowing out the state does not exclude the machineries of coercion. Traditional security studies tends to focus on the state as both the key referent object and the holder of legitimate right to use force, and to see all military affairs as security. But in a liberalized world the logic of privatization moves into the realm of coercion both domestically and internationally. Domestically, the hollowed-out liberal Leviathan is abandoning its monopoly on the use of force: in the US, Britain, Canada and Australia private security personnel outnumber public police by two or three to one; in Russian and South Africa, by ten to one. 97 Many companies and "gated communities" provide much of their own policing. Within all types of state, substate and transnational groups are claiming and exercising the right of self-defence. Internationally, most notably in Africa, mercenary firms such as Executive Outcomes are becoming significant transnational military players. It is far from clear where this trend towards the privatisiation of violence leads. In liberal economic logic it makes quite good sense to avoid the expense of maintaining a standing army, and just to hire one when you need it. But the political consequences of this are quite profound, going right to the core of what the modern state is supposed to be about. So also are the consequences of privatised domestic security, which not only opens up questions of uneven provision of security for citizens, but also questions about political accountability and democratic control.

The other side of privatised violence is transnational criminal mafias, who benefit from the openness of liberal systems, and constitute one of the dark sides of increased openness. As national borders become more open, the opportunities for transnational crime become greater. As a rule, it is far easier for mafias to extend their activities across borders than it is either for national police forces to upgrade their cooperation to comparable levels, or for states to agree to set up international police forces. Even within the EU, where interstate cooperation is quite highly developed, this is a problem. Transnational crime in many ways resembles trade (in drugs, weapons, migrants, stolen cars, and suchlike), which is why it benefits from a liberal economic order. But some of the mafias have acquired significant national and transnational coercive capability, sufficient to outgun the local police and army in some states.

In part because of the blinding success of the classical liberal project in solving the traditional military problem, not much coherent thought has yet been given to the question raised by these issues about the military character of a liberalized world. 98 One issue does, however, stand out very clearly, and that is the tension between liberal commitments to reduce restraints on trade on the one hand, and military security concerns about the proliferation of both light weapons and weapons of mass destruction on the other. This bears some resemblance to the facilitation of organised crime by liberal rules. The openness of liberal systems facilitates trade in militarily significant technologies, the moreso as states and firms come under relentless pressure to compete for markets by selling whatever they can. This pressure runs up against fears in the zone of peace that "rogue states" in the zone of conflict will acquire weapons capable of enabling them either to dominate their regions, or to threaten the zone of peace. The stories of how Iraq and Pakistan acquired technology for making nuclear weapons illustrate both the commercial pressures and the highly problematic interweaving of civil and military technologies in many areas (so-called "dual-use"). 99 As well as pointing to one of the key military security issues spilling over from liberal economic praxis, these cases also point to a deeper problem.

Given that industrialization inevitably carries with it increasing capabilities for military production, how are liberals to deal with unliberal, revisionist, modernist states such as Iran, Iraq and China? To pursue trade and investment with them is to gamble that the liberal logic of interdependence and domestic transformation (from market to democracy) will work more quickly and powerfully than the realist logic of foolishly strengthening an opponent that you may one day have to fight. The inconsistencies of American policy towards China and Iran illustrate the difficulties of choice, and perhaps nowhere does this liberal-realist dilemma operate more clearly than in relation to a potential superpower such as China. By engaging with the Chinese economy Asian and Western traders and investors enrich themselves and China, and entangle Beijing in the liberal incentive scheme of joint gains requiring peaceful relations. But by enriching a still authoritarian China, and upgrading its technological capacity and economic weight, they also make it more powerful, increasing its means to make trouble should its leaders want to go in that direction. There is no easy escape from this dilemma. Traders and investors are competing with each other, making coordinated action difficult. If Japan, for example, began to fear Chinese power it could not easily switch from liberal to realist behaviour. Doing so would simply turn the profits from the Chinese market over to other players, weakening Japan, and antagonising China. Taiwan already faces this dilemma in acute form. So long as the liberal logic remains strong, as it seems certain to do in the zone of peace, China will be able to feed on the resources of those it may later wish to confront. Only if its behaviour managed to frighten all of its neighbours, as well as the West, would there be any possibility of coordinated economic action to restrain China's power.

6. Conclusion: Implications for Security Policy and Security Analysis

The extent of the security agenda unfolded in the previous section, and the depth of its linkages to "real existing liberalism", suggest that the undoubted success of the classical liberal project has by no means closed the book on security. To assume that the post-Cold War world has been successfully desecuritised, or that only military security issues remain, risks a major misunderstanding. Although it has made enormous gains against the classical insecurity of war, liberal economic and political praxis is itself now shaping and generating a reopening of security across a wide agenda. What we are witnessing is not, as touted by liberal triumphalists, the closure of a single problematique about government, economy and security. Rather, we seem to be looking at a more dialectical process in which the resolution of one phase opens up a new configuration with a new set of tensions.

The exact nature of this new phase is far from clear, but Cerny perhaps comes close to capturing it with his idea of a growing disjuncture between state level political institutions largely shaped by the "political economies of scale" of the second industrial revolution, and transnational economic organizations increasingly shaped by those of the third. He sees this disjuncture as undermining the ability of the state to provide either collective goods or legitimate democratic accountability. 100 In this view, the close linkage between political and economic institutions (i.e. the modern state) during the second industrial revolution underpinned the familiar inter-state security agenda that classical liberalism set out to solve. The increasing divergence in the political economies of scale driven by the third industrial revolution (i.e. the postmodern, or what Cerny calls the "competition" state), similarly underpins the wider security agenda outlined above. This new configuration opens up a whole range of security issues many of which cannot even be addressed, let alone solved, within the existing division of labour between security studies and IPE.

What can be clearly observed is that the state is less important in the new security agenda than in the old one. A range of new referent objects for security and sources of threat is being set up above, below and alongside the state. Above the state as new referent objects one finds the set of rules, regimes and institutions that constitute the liberal international economic order (LIEO); the global climate system; and the various regimes that attempt to control the proliferation of weapons of mass destruction. 101 Alongside the state, nations and religions have emerged as distinct referent objects. 102 Below it, the rising focus on human rights supports claims to give individuals more standing as the ultimate referent object for security. 103 At the same time, the sources of threat are also diversifying away from the state. Many of the new threats seem to stem from complex systems both natural (the ecosystem) and human-made (the global economy), and the operation and interplay of these systems is often poorly understood.

What this development points to is a powerful case for dissolving most of the institutional and attitudinal barriers between security studies and IPE. These two sub-disciplines cannot any longer assume that they are dealing with largely separate, or even just analytically separable, realms. Such a dissolving of boundaries will be difficult for both sides, and not just because of institutional inertia. Although they are keen to dissolve state boundaries, liberals have been powerful promoters of the politico-conceptual separation of the economic and political sectors. Much of security studies rests on a belief in the ultimate primacy of military power, and on the intellectual cozyness of a tightly defined community of expertise in military matters. IPE, as its name implies, is already a rejection of the liberal premise of sectoral separation between economics and politics, but has nevertheless grown up in part by defining itself in contrast to security studies. Nothing being proposed here prevents the pursuit of specialization. But it does require that such pursuit occur within a wider framework of awareness about the overall character of the "political economies of scale" within which events occur

A degree of integration between IPE and security studies would open up at least three opportunities.

First, it would help to break down the now outdated conceptualization of security that it is about objective threats defined predominantly in military terms. It would speed the replacement of this view with the idea that security is about socially constructed threats, having an existential quality and leading to a call for emergency measures. As we have argued elsewhere, such a reconceptualization both addresses the wider agenda, and solves the problem of incoherence that unlimited widening poses to the concept of security. 104

Second, it would liberate security theory from the state-centric, inter-state relations models that now dominate it. While such purely political models retain considerable use in the post-Cold War world, they both fail to capture, and restrict the view of, the redefinition of political economies of scale that underpins much of the contemporary security agenda. More integration of expertise about the politics of threat on the one hand, and the interplay of authority and markets on the other, would unfold a more salient and more fruitful, if also more complex, theoretical foundation for security studies. And given the close linkage between security studies and realism, this new synthesis might also enliven thinking about general theories of the international system. It is not without significance that arguably the most successful of these theories, neorealism, is rooted in an old-style, state- and military-based conceptualization of security.

Third, it would put some useful pressure on liberals to start seeing their doctrines as part of the problem, and not just as a definition of all things good and progressive. It might also make them abandon the myth that politics and economics can be separated for either analytical or practical purposes. It would force them to confront the political and security consequences of liberal praxis, and to place urgently on their agenda the need to consider how to deal with the threat of resecuritisation that the wider agenda poses to the desecuritising goals of the classical liberal agenda. The (re)linkage of security studies and IPE brings into sharper perspective some crucial questions. One, as Williams argues, is that the current resecuritising of identity threatens some of the foundational moves of liberal thinking that sought precisely to find a way out of the violence and insecurity of the European religious wars. 105 Another, is the apparent threat posed to democracy by the effect of liberal economic praxis on the state. This threat attacks not only the heart of the whole liberal ideological construct, but also the main foundation of the liberal theory of democratic peace.

This need for a more integrated understanding of the new security is not just a matter for academics. Coherent security policy also depends on an ability to see clearly the linkages between how the political economy works, and what gets constructed as a security issue. The much observed disjuncture between US security policy (trying to be global hegemon) and economic policy (seeking as much immediate advantage as possible), is only one example of what happens when these linkages are not made or not understood. 106 Another is how the integration project of the EU has run up against societal security problems in several of its constituent nations. The triumph of liberalism has dissolved many of the traditional security problems. But it has at the same time generated a new and in some ways unfamiliar security agenda. Both academics and policy makers need new ways of thinking if they are going to meet this challenge successfully.



Note 1: Arblaster 1984; Bramsted & Melhuish 1978; Dunn 1993; Gray 1986; Hartz 1955; Laski 1936; Manning 1976; Morgenthau 1946; Richardson 1997; de Ruggiero 1927; Smith 1992; Back.

Note 2: Cf Lakoff 1996 Back.

Note 3: This approach is set out in Wæver, 1995; and Buzan, et al, 1997, esp. ch. 2. Melvyn P. Leffler, too, has argued that the (since Wolfers 1952 famous) ambiguity of "security" is a problem to policy makers, but to the historian an advantage because it allows the study of the varying definitions of threats, core values and legitimate security interests (1990: 145). For a similar argument see Jahn et al 1987, pp.. On constructuvist approaches to security, see also Katzenstein, 1996. Back.

Note 4: Hobbes (1968 1651), Tuck 1984; Sorrell 1986; Williams 1996, forthcoming Back.

Note 5: Hobbes (1968 1651), ch. 21 / p. 264. Back.

Note 6: Sorrell 1986: 121ff Back.

Note 7: Hobbes 1968 1651), ch. 21, p. 270. Back.

Note 8: Williams 1996; Tuck 1984; Hirschman 1977; Bartelson 1993 pp Back.

Note 9: See documentation and quotes in Rotschild 1995 Back.

Note 10: Ibid., p. 63. Back.

Note 11: Eg. Locke 1924 1689, 123f, 159f, 183-186 pp pp; Bramsted & Melhuish 1978, 106f, 245, pp; Montesquieu 1989 1748, pp, Mill 1989 1859, pp; Kant 1784, fifth, sixth and seventh "sentence". Back.

Note 12: Gordon 1991: 33; cf. Foucault 1991 [1978], 1997; Boldt 1972; Locke 1924 1689, 188, 197-203 Back.

Note 13: Boldt 1972; cf. also Luhmann 1989 1980: 133-4. Back.

Note 14: Cf. Wæver 1995, xx ; Schmitt 1934 [1922]. Back.

Note 15: Gordon 1991; Foucault 1991 1978, 1997; Burchell 1991; Back.

Note 16: Gordon 1991; Foucault 1991 1978, 1997; Luhmann 1989 1980; Pasquino 1991; Skinner 1989 Back.

Note 17: Polanyi 1957 1944 Back.

Note 18: Gordon 1991; Burchell 1991, 139; Conze 1984, 849. Back.

Note 19: Bramsted & Melhuish 1978, 21. For similar identification of liberty and security, see Wilhelm von Humboldt quoted in Conze 1984, 851f and in Bramsted & Melhuish 1978, 272-4; and Kant quoted by Conze 1984, 851. Back.

Note 20: Burchell 1991: 143ff; see also Foucault 1991, 1997. Back.

Note 21: Shklar 1984: 244, cf. also Shklar 1989 Back.

Note 22: Morgenthau 1946; Ashley 1987; Hurrell, 1995: 455-7 Back.

Note 23: Rousseau in Hoffmann & Fidler 1991, xvii, xix-xx; Bailyn 1992 [1967]; Tocqueville 1959 1835; Earle, 1970 [1943]: ch. ?; Lasswell 1950; Back.

Note 24: Kant 1784; Bramsted & Melhuish 1978, 278-287; Keohane 1990; Moravcsik 1993; Smith 1992; Morgenthau 1946; Carr 1939; Zacher & Matthew 1995; Hoffmann & Fidler 1991 Back.

Note 25: Smith 1992, 206 Back.

Note 26: Zacher and Matthew, 1995: 111-20 Back.

Note 27: Claude 1984 1956; Carr 1939 Back.

Note 28: Norman Angell, David Mitrany, etc - see the re-readings of especially de Wilde 1991 and Navari 1989. Back.

Note 29: Quoted in Boll 1988: 57. Back.

Note 30: 1 Lippmann 1943; Morgenthau 1946; Wolfers 1956; Huntington 1985 [1957]: 148-150. Back.

Note 31: Bailyn 1992 [1967]. Back.

Note 32: On the pre-1945 history of the word security (mostly in other contexts than "national security"), see Halpérin 1952, Febvre 1956, Kaufman 1970, Conze 1984; Delumeau 1986; Der Derian 1993, Dillon 1995. Back.

Note 33: Yergin 1977: 196, 220, 450; cf. also Mandelbaum & Yergin 1973. Back.

Note 34: Yergin 1977: 195 Back.

Note 35: Rosenberg 1993: 279-81. Back.

Note 36: Lippmann 1943, Back.

Note 37: Hamilton et al ,1911 1787-8; Bailyn 1992 [1967]; Deudney 1995. Back.

Note 38: Etzold 1978, pp. 1-2. Back.

Note 39: Leffler 1980, 1984, 1990, 1992, 1995. For a partially diverging interpretation, see Gaddis 1982 Back.

Note 40: Leffler 1992, pp; Buzan 1984, 604f. Back.

Note 41: Lasswell 1950, p. 22. Cf. Lasswell on the "garrison-police state", 1950: 47f, 1941, 1962. Back.

Note 42: Huntington 1985 [1957]: p. 456. Back.

Note 43: Truman 1952-53; quoted by Leffler 1992, p. 13. Back.

Note 44: Yergin 1977, 302; Leffler 1992; Hogan 1987 Back.

Note 45: Leffler 1992, 343-4; Arblaster 1984, 313-6; Campbell 1992: 166-187 Back.

Note 46: Gardner, 1980:7-9. Back.

Note 47: Ruggie 1982: 393 Back.

Note 48: Hoffmann 1966 Back.

Note 49: Buzan, 1987; chs. 2,11; Freedman, 1989 [1982]. Back.

Note 50: Glaser, 1984 Back.

Note 51: McKinlay and Little 1986, 172-80 Back.

Note 52: Arblaster 1984, 299-308 Back.

Note 53: Hirshman 1977; Williams forthcoming Back.

Note 54: Ibid., pp. 307f. Back.

Note 55: Arblaster 1984, 308. Back.

Note 56: Arblaster 1984, 317-32. Back.

Note 57: Allison et al., 1985 Back.

Note 58: Matthews 1989; Homer-Dixon 1991; Buzan, Wæver and de Wilde 1997, ch. 4. Back.

Note 59: Walt, 1991. Back.

Note 60: Gray 1994b; Chipman 1992: 129. Back.

Note 61: Walt 1991, 212-3 Back.

Note 62: Walt 1991 227; see also Dorff 1994; Gray 1994a Back.

Note 63: This line of argumentation was suggested to us by Michael C. Williams as a way of extending our analysis in Buzan et al 1997; see Williams forthcoming. Back.

Note 64: Deudney 1990 Back.

Note 65: Buzan 1995, Wæver 1995 Back.

Note 66: Doyle, 1995; Hurrell and Woods, 1995. Back.

Note 67: Fukuyama 1992. Back.

Note 68: Huntington 1993, 1996 Back.

Note 69: Moravcsik 1993; Owen 1994 Back.

Note 70: Brown 1992; Latham 1993, 1995, 1996; Gordon 1991: 18; Burchell 1991: 143; Foucault 1997, 73-79 Back.

Note 71: A general overview of the nature of security in the different sectors is found in Buzan et al 1997,ch. 3-7. The text below draws on this in addressing the specific issue of global liberalism as the cause of securitisation. Back.

Note 72: Luciani, 1989; Buzan, 1991b, ch. 6; Cable, 1995; Buzan, Wæver and de Wilde, 1997: ch. 5. Back.

Note 73: Ruggie 1995; cf. also Ruggie 1982 and1991 (and 1994?) Back.

Note 74: Paye,1994 Back.

Note 75: Strange, 1986 Back.

Note 76: Wallerstein, 1993; Huntington, 1993, 1996; Spence, 1994, p. 4; Kaplan, 1994 Back.

Note 77: Rosenau 1990; Ruggie 1993; Cerny 1995 a or b?; Strange 1994. Back.

Note 78: Richardson 1997, 16. Back.

Note 79: Latham 1996; and Bramsted & Melhuish 1978, 37-9, on the rise of national-liberalism in the 19th Century. Back.

Note 80: Wæver et al., 1993 Back.

Note 81: Siklová 1991 + globalization sociologists? Back.

Note 82: Hassner 1993, 1996; for the contrasting expectations, see eg. Habermas 1992 Back.

Note 83: Cf. von Hayek 1944 Back.

Note 84: Buzan and Segal, 1996, 1997 ch. 6; Holm & Sørensen 1995; Caporoso 1996 Back.

Note 85: Booth, 1991; Shaw, 1993. Back.

Note 86: Soros, 1997 Back.

Note 87: Polanyi 1957 1944, ch. 19 Back.

Note 88: Ripsman and Blanchard, 1996-7.? Back.

Note 89: Buzan and Segal 1996; Luttwak 199?; Strategic Survey 1996-7 pp; Latham 1996: 84 Back.

Note 90: Hirsch and Doyle 1977; Keohane 1980 and 1984; Strange 1984; Gilpin 1987; Kindleberger 1981. Back.

Note 91: Strange 1994 Back.

Note 92: Keohane 1984. Back.

Note 93: Gilpin, 1996: 3. Back.

Note 94: Wallerstein, 1974:390-412; Galtung, 1971: 85-91; Calleo and Rowland, 1973:11; Gill and Law, 1988: 98-9 Back.

Note 95: Buzan, 1991a, p. 432; Goldgeier and McFaul, 1992; Singer and Wildavsky, 1993; and implicitly in an earlier version Keohane and Nye, 1977. Back.

Note 96: Luttwak ... Back.

Note 97: Economist, 19 April 1997:23 Back.

Note 98: One exception is Cerny 1995b, 50-53 Back.

Note 99: Dallmeyer, 1995; Kapur, 1987 Back.

Note 100: Cerny, 1995a Back.

Note 101: Buzan, Wæver and de Wilde, 1997: chs. 3-8. Back.

Note 102: Wæver, et al. 1993 Back.

Note 103: Shaw, 1993; McSweeney, 1996; Booth 1991, 1994, 1995; Rotschild 1995 Back.

Note 104: Wæver 1995; Buzan and Wæver, 1997; Buzan, Wæver and de Wilde, 1997 Back.

Note 105: Williams, 1977. Back.

Note 106: Mastanduno, 1997, pp. 28-9. Back.