The Democratic Peace

The Democratic Peace is grounded in the empirical observation that democracies rarely if ever go to war against one another. Besides this, for some authors it has come to encompass a wider range of empirical phenomena as summarized by Bruce Bueno de Mesquita: democracies will fight non-democracies; they tend to win a disproportionate share of their wars; when disputes emerge between democracies they are more likely to be settled peacefully than when between other pairings; democracies are more likely to initiate wars against autocracies than vice versa; those wars they do initiate are likely to be shorter than the wars of non-democracies and result in fewer lives lost; transitional democracies appear more likely than established democracies to fight wars; and larger democracies seem more constrained than do smaller democracies to avoid war [1999, 791; but see, for example, Russett and Oneal (Reference Russett and Oneal2001) and McFaul (Reference McFaul2007) regarding the war-proneness of transition democracies]. It is, however, the proposition of peace between democracies that remains the cornerstone of the research programme (for reviews of the literature see Chan Reference Chan1997; Ray Reference Ray1998; MacMillan Reference MacMillan2003; Rosato Reference Rosato2003; Gleditsch Reference Gleditsch2008; Lektzian and Souva Reference Lektzian and Souva2009; Geis and Wagner Reference Geis and Wagner2011).

Michael Doyle's seminal 1983 essays grounded the observation in Kant's (Reference Kant1795) essay, Perpetual Peace, to argue that the expansion of liberal republics in conjunction with the establishment of an international confederation and cosmopolitan exchange had formed the bases of an enduring and potentially universal peace. Indeed, through charting the future development of the international system according to two tracks, one being democratization, trade, and cultural exchange and the other the incidence of war, Doyle calculated that ‘global peace’ should be established sometime in the first decade or two of the 22nd century (1983b, 352). One valuable feature of Doyle's work that has sometimes been overlooked is that his account is concordant with arguably the single most important insight of the liberal internationalist tradition: that a stable peace must be grounded in justice. For Doyle, ‘domestically just republics, which rest on consent, presume foreign republics to be also consensual, just, and therefore deserving of accommodation’ (1983a, 230). Justice, for Doyle, arises from the development of law-governed communities in which individuals are treated as ends in themselves through some combination of civil and political rights, social and economic rights, and the right of democratic participation as a guarantee of the first two (1983a, 206–07). Doyle's republics, then, are conceptualized as ‘thick’ moral–political communities comprising meaningful sets of democratic rights.

With the end of the Cold War, Bruce Russett's Grasping the Democratic Peace:Principles for a Post-Cold War World argued that the expansion of democratic regimes in the 1990s was an opportunity to transform international politics through replacing the ‘war system’ established by autocracies centuries earlier with the peaceful practices of negotiation and compromise associated with the domestic behaviour of democracies (1993, 137–38). The Kantian theme re-emerges in Triangulating Peace (2001) written with John Oneal in which the authors generate empirical evidence in support of the ‘virtuous circles’ that can emerge from democracy, trade and membership of international organizations such as to form a ‘Kantian Peace’. They argue that these factors serve to progressively reform and pacify the international system, in contradistinction to the emphasis upon conflict central to rival visions of the future such as the ‘clash of civilizations’ thesis. In all of these works, the authors go to considerable lengths to caution against ‘liberal crusading’ and emphasize the need for liberal Powers to exercise ‘moral authority’ and behave as good international citizens. But there is little question that running through these central accounts of the Democratic Peace is an explicit or implicit liberal teleology in which the expansion of democracy, trade, and (to a lesser extent) international organizations are the key to peace. Indeed, it is the ‘promise’ that the spread of democracy will see the withering of war that gives the research programme its rhetorical force.

The most fundamental division within the literature has been between the so-termed ‘dyadic’ and ‘monadic’ positions referring, respectively, to whether democracies are peaceful only in relation to one another or in general. The dyadic is generally favoured (Doyle Reference Doyle1983a, Reference Doyleb; Russett Reference Russett1993; Owen Reference Owen1997; Bueno de Mesquita et al. Reference Bueno de Mesquita, Morrow, Siverson and Smith1999; Russett and Oneal Reference Russett and Oneal2001), but the monadic position retains some support (Benoit Reference Benoit1996; Rummel Reference Rummel1997) and even some dyadic writers regard it as of empirical interest (Russett and Oneal Reference Russett and Oneal2001). The starkness of the dyadic/monadic divide has been challenged by MacMillan (Reference MacMillan2004) who applied an understanding of liberal norms derived from liberal international theory and a more critical account of the liberal state to identify a more nuanced pattern of Democratic Peace- and war-proneness. In what promises to be a major contribution, Geis et al. Reference Geis, Müller and Schörnig(2013) identify specific patterns of war involvement by individual democracies and explain variation in war involvement between democracies through a constructivist account that ties conflict behaviour to respective notions of national identity and the legitimacy of force. Nevertheless, the monadic–dyadic distinction is important as it has implications for the task of theorizing the phenomenon. Proponents of a dyadic peace not only have to give an account of inter-democratic state peace but also why this does not extend to relations with non-democracies. This requirement is apparent in normative and institutional (Maoz and Russett Reference Maoz and Russett1993), informational (Schultz Reference Schultz2001), and rationalist institutional arguments (Bueno de Mesquita et al. Reference Bueno de Mesquita, Morrow, Siverson and Smith1999; Bueno de Mesquita et al. Reference Bueno de Mesquita, Smith, Siverson and Morrow2003).

Whilst it is not possible to review all of the dyadic explanations that have emerged in the literature, Maoz and Russett's (Reference Maoz and Russett1993) influential early account is indicative of the way in which these accounts tend to deflect any notion that responsibility for the violence of democracies may be due to some factor internal to the liberal or democratic state. Instead, such responsibility is typically situated in the nature of the non-democratic state, the strategic interaction process, or the need to adjust one's behaviour in line with the Hobbesian geoculture that persists beyond the pacific union. For Maoz and Russett democracies externalize their domestic norms of cooperation and conflict such that when two democracies confront one another these traits flow through to the process of conflict resolution internationally. But ‘when a democratic state confronts a nondemocratic one, it may be forced to adapt to the norms of international conduct of the latter lest it be exploited or eliminated by the nondemocratic state that takes advantage of the inherent moderation of democracies’ (1993, 625). The structural (or institutional) argument rests on the need for states to mobilize key groups in support of a conflict. The complex political structure of a democracy in which governments need the support of public opinion and a range of institutions narrows the range of issues over which a democracy is likely to fight and makes mobilization difficult and cumbersome. Thus, should a conflict arise between two democracies then there is greater time for a negotiation and compromise. In a conflict with a non-democracy, however, the pace of dispute is set by the relatively less-constrained non-democracy such that democratic leaders ‘are forced to find ways to circumvent due political process’ (1993, 626).

That, given the historical record of imperialism, conquest, and war initiated by democracies such explanations of democratic state violence were obviously incomplete, was a point not lost on Bueno de Mesquita et al. (Reference Bueno de Mesquita, Morrow, Siverson and Smith1999, Reference Bueno de Mesquita, Smith, Siverson and Morrow2003) who acknowledged that strong democracies sometimes initiated conflict against weak states including against other democracies. From the foundational assumption that political leaders seek to retain office, they argue that the strategies required to achieve this are a function of institutional context and hence are regime-dependent. In an autocracy, leaders need only to satisfy a relatively narrow ‘selectorate’ in order to retain office and hence are able to do this through the distribution of ‘private goods’. By contrast, in a democracy the breadth of the selectorate means that the key to retaining office is success in the delivery of public policy. From this, peace between democracies is maintained because a democracy will be wary of fighting another democracy because it recognizes that it too will ‘try hard’ and commit a high proportion of its resources to the conflict. Hence, any war between them will entail higher cost and higher risk that if they were fighting an autocracy.

This strictly rationalist approach carries with it a sting in the tail for the Democratic Peace, however, in that democracies will not be so constrained in the use of force against a weak state, including a weak democracy, when victory is likely to be assured and leadership survival is not put at risk (Bueno de Mesquita et al. Reference Bueno de Mesquita, Smith, Siverson and Morrow2003). But, this simple cost–benefit calculation does not help one understand those instances when a military victory would be perfectly possible but yet there is no use of force, and is potentially at a disadvantage when set against theoretical accounts that can identify and explain more specific patterns or distributions of democratic state violence (including that between democracies), as do certain strands of norm-based theory discussed above and by world-systems theory below. The attention to violence between strong and weak democracies such as the US attack on the Dominican Republic in 1965 does, however, highlight where the Democratic Peace frays at its edges (see also discussion below), which would also appear to be the case with the effects of commerce and interdependence.

There is a significant body of empirical work that supports the notion of a positive relationship between trade, interdependence, and peace, reinforcing the point made by Norman Angell and others that trade interdependence diminishes the rationality of war (Angell Reference Angell1913; Rosecrance Reference Rosecrance1986; Chan Reference Chan1997, 76; Oneal and Russett Reference Oneal and Russett1997; Russett and Oneal Reference Russett and Oneal2001, 125–55; Mousseau et al. Reference Mousseau, Hegre and Oneal2003, 279–80; Schneider et al. Reference Schneider, Barbieri and Gleditsch2003; Gelpi and Grieco Reference Gelpi and Grieco2008; but see Barbieri Reference Barbieri2002). However, several authors argue that the pacific benefits of trade are limited to relations between advanced capitalist states such that this aspect of the Democratic Peace is more a function of development than democracy.

Mousseau et al. (Reference Mousseau, Hegre and Oneal2003) argue that trade and commerce become insignificant indicators of peace if at least one party in a dyad falls below a certain level of development (see also Hegre Reference Hegre2000, 6; Mousseau Reference Mousseau2005). For Mousseau, ‘it is thousands of times more likely that the democratic peace is a phenomenon limited to nations with above-median levels of development than it is that democracy impacts all nations equally, regardless of wealth’ (2005, 73). Whilst, then, relations between developed democracies may still be regarded as especially peaceful, interactions between rich and poor democracies ‘occur as they do between non-democracies’ such that ‘coercion is a tool for settling differences’ (Mousseau et al. Reference Mousseau, Hegre and Oneal2003, 286; but see Russett Reference Russett2010, 201). These empirical claims have been matched at the theoretical level by the emergence of the ‘capitalist peace’ literature in which ‘democracy’ has been jettisoned in favour of the claim that the Peace is a product of economic development, similar interests and the globalization of capital (see Weede Reference Weede2003; Gartzke Reference Gartzke2007; Schneider and Gleditsch Reference Schneider and Gleditsch2010; but see Choi Reference Choi2011).

Following this brief account of the Democratic Peace and its limits, the discussion turns now to the critical materialist tradition. Of particular importance are the account of the nature of the world-system and the relationship between the state and capital, notions of core and periphery in the hierarchical world economy, and the concept of the social structure of accumulation. But besides this is the epistemological point that informs how one might think about the construction of the Democratic Peace and the stability of its key conceptual referents. For Wallerstein, ‘when one is dealing with a complex, continuously evolving, large-scale historical system, concepts that are used as shorthand descriptions for structural patterns are only useful to the degree one clearly lays out their purpose, circumscribes their applicability, and specifies the theoretical framework they presuppose and advance’ (1983, 100). This important insight informs discussions of ‘the state’, ‘the international system’, ‘democracy’, and ‘war’ below.

Critical materialism and the Democratic Peace

There has been rather little direct attention to the Democratic Peace research programme from within the critical materialist tradition and what there has been largely ignored, albeit not without the occasional pot shot from the liberal camp (see Gleditsch Reference Gleditsch2008, 703). From a world-systems perspective, Christopher Chase-Dunn and Bruce Podobnik (Reference Chase-Dunn and Podobnik1999) briefly discussed the Democratic Peace and whilst they regarded it as empirically valid in its own terms they were sceptical about whether it would be a permanent feature of international relations. It is, however, Barkawi and Laffey (Reference Barkawi and Laffey1999, Reference Barkawi2001) and Barkawi (Reference Barkawi2001) that provide by far the most extensive critical perspective on the Democratic Peace and the discussion below has benefited from their work.

Critical materialism comprises a broad tradition centrally concerned with the problems the historical development of capitalism and capitalist social relations have presented for human freedom. Whilst the discussion draws on a range of insights from the tradition, it is grounded principally in world-system theory, which is used as a platform through which to engage and critique the Democratic Peace. Accordingly, the argument is developed through the work of self-identified world-system theorists (such as Immanuel Wallerstein, Christopher Chase-Dunn, Giovanni Arrighi, and Samir Amin); those who explicitly acknowledge the influence of world-system theory on their work but who have sought to go beyond or supplement it in some way (such as William Robinson's complementary use of world-system and Gramscian approaches); and those whose work bears directly upon key world-system concepts (such as the comparative political sociologists Rueschemeyer, Stephens, and Stephens) for their analysis of the dynamics and implications of ‘dependency’ for political development and democratization.

World-system theory (and dependency theory from which it drew) departed from Marx's own position on the prospects for development under capitalism beyond the core of industrially and technologically advanced nations. Marx clearly indicated in the Communist Manifesto and his writings on India that he expected capitalism to develop the periphery. By contrast, Wallerstein and others such as Frank (Reference Frank1993) have argued that since the expansion of the world-system from the 16th century, but particularly from its increased expansion across the globe in the 18th and 19th centuries, the trend has been for the peripheral and semi-peripheral areas to become trapped in the production of raw materials and semi-manufactured goods for export onto the world market and into the importation (on unfavourable terms) of manufactured and technologically advanced goods from the core (Wallerstein, Reference Wallerstein1989). This ‘classic’ view was, however, upset by the rise in the 1980s/90s of the East Asian economies led by Japan and more recently of China and the BRIC economies, whilst variation in the rate of growth within the periphery and semi-periphery reflected the fragmentation of the rigid geographical separation of core and periphery and showed that structural factors are not necessarily an impediment to economic development and democratization under capitalism.

The place of world-system theory within the wider critical materialist tradition is controversial, most notably for its grounding not in the primacy of class relations but in a world economic system characterized by processes of unequal exchange, occurring through a series of commodity chains spanning core and periphery zones in which the capital intensive zones have a stronger bargaining position by approximating monopoly (or oligopoly) conditions in contrast to the more competitive labour-intensive production of the periphery (Brenner Reference Brenner1977). There are, however, three key reasons why it is valuable as a starting point for analysis of the Democratic Peace. First, its attention to states helps to forge the necessary bridge between class- and state-based analyses that presents genuine problems for establishing terms of engagement between the critical materialist and liberal traditions. As Linklater has noted, the world-system approach sets out ‘to recover the state, international relations, and war for modern social theory’ (Reference Linklater1990, 119). The role of the state and in particular the modern states-system in sustaining the capitalist system of production is discussed below.

Second, the bold world-system categories of ‘core’ and ‘periphery’ may be regarded as analytical counterpoints to the equally bold division within Democratic Peace theory between ‘democratic’ and ‘non-democratic’ regimes. Significantly, whereas Democratic Peace theory has tended to regard the democratic and non-democratic realms as having developed along different and for all effective purposes unconnected lines, world-system theorists regard the development of core and periphery as intimately connected and to some extent mutually constitutive. To quote Wallerstein,

Wallerstein's point that core and periphery are primarily ‘relational’ terms is employed below to highlight the shifting geography or spatiality of the core-periphery interface under conditions of globalization and following from this the changing contours of core-periphery conflict and violence, which are not confined to the inter-state analysis typical of the Democratic Peace.

Third, whereas the Democratic Peace seeks to explain the effects of regime type on the war-proneness of states, world-system theory purports to identify a set of social forces that are drivers of both regime type and war. Hence, for example, Chase-Dunn and Podobnik's reservations upon the permanence of the Democratic Peace arises from concern that the tensions associated with underlying economic and hegemonic cycles will create a situation in which one or more democracies will revert to authoritarianism, thereby increasing the probability of war (1999, 53; see also Wallerstein Reference Wallerstein1983). They identify in particular an anticipated ‘window of vulnerability’ in the 2020s in which the late phase of a Kondratieff cycle upswing occurs simultaneously with greater systemic multipolarity and an intensification of rivalry between core powers for raw materials and markets (1999, 45). Whilst it is fair to say that world-system theorists differ on the nature of the mechanisms that produce war (1999, 46–47), the notion that economic factors might influence both regime type and the probability of war adds a prospectively promising layer of analysis.

But how does critical materialism explain the empirical phenomenon at the heart of the Democratic Peace? The most widespread explanation of the absence of war between (mature) democracies is as the expression of common capitalist interests in the domination of the periphery in the vein of Kautsky's notion of ‘ultra-imperialism’. Here, ‘far sighted’ capitalists recognized that the costs of colonial expansion and imperialism threatened the rate of capital accumulation and generated arms races and war, leading them instead to pursue the ‘cartellization’ of foreign policy through a ‘holy alliance of the imperialists’, united on the one hand against the agrarian, peripheral zones and on the other their own domestic proletariat (Kautsky Reference Kautsky1914; see also Arrighi 2010, 39).

Kautsky's position differs from Lenin's (Reference Kautsky1917; see also Brewer Reference Brewer1990) influential view that the contradictions of capitalism led directly to imperialism and to war. Indeed, whilst the role played by capitalist development and modernization in the outbreak of the three world wars of the 20th century – two hot and one cold – is still debated (see e.g. Berghahn Reference Berghahn1993; Neocleous Reference Neocleous1997; Halliday Reference Halliday2010), what is clear is that since 1945 relations between the advanced capitalist states have been marked more by cooperation than by conflict. In this vein, Barkawi and Laffey regard the liberal Democratic Peace as a product of being ‘embedded in geostrategic and political economic relations that buttress international state and capitalist power in hegemonic, i.e. non-violent, ways’ (1999, 419). Christopher Chase-Dunn and Bruce Podobnik acknowledge the view that ‘unrivalled US military dominance, the strengthening of international institutions, and the continuing consolidation of a unified world capitalist class are all seen as promoting peace and stability within the core of the world-system’ (Reference Chase-Dunn and Podobnik1999, 40) but as noted above regard the phenomenon as historically contingent upon wider trends in the world economy.

For world-system writers, the world-system, comprising the inter-state system and the world economy, is structurally crisis-prone but has evolved through a series of hegemonic cycles in which successive hegemons have overseen the development of an increasingly extensive set of functional capacities in response to the system's increasing complexity. The world-system is reproduced through a mutually constitutive relationship between the modern state and capitalism in which capital provides the resource base for state consolidation and success in the wider geopolitical competition between states whilst the state assists in ensuring the reproduction of the conditions for capitalist accumulation. In particular, this entails the maintenance of a ‘partially free’ market wherein the division of the world into distinct juridical domains in the form of ‘the interstate system provides the requisite political framework to facilitate the manoeuvrability of capital such as to escape organized workers and other social constraints on profitable accumulation’ (Chase-Dunn Reference Chase-Dunn1999, 211; see Arrighi 2010, 33). For Wallerstein, ‘capitalism and the modern state-system were not two separate historical inventions… [but] simultaneously developed, and neither could continue to exist without the other’ (Wallerstein Reference Wallerstein1996, 89).

However, whilst the modern state and the rise of the capitalist world economy are regarded as co-constitutive and interdependent, they exist in relations of dynamic tension rooted in their respective bases of power. The ‘territorialist’ logic of state-power stresses the size of population and territory as the key denominator, with the pursuit of wealth regarded as a means to this end, whereas the the ‘capitalist’ logic of power emphasizes command of scarce resources, with territory and population valued in so far as they are instrumental to this end. For Arrighi, ‘the critical feature of the [modern interstate] system has been the constant opposition of the capitalist and territorial logics of power and the recurrent resolution of their contradictions through the reorganization of world political-economic space by the leading capitalist state of the epoch’ (2010, 37). In contradistinction to the liberal account, historically periods of discontinuous change – crisis, restructuring, and reorganization – have been far more common in the evolution of the capitalist world economy than ‘those brief moments of generalized expansion along a definite developmental path like the one that occurred in the 1950s and 1960s’ (Arrighi 2010, 1).

World-system theorists do not deny that the state pursues the type of geopolitical competition at the centre of realist accounts of international politics, but rather emphasize the connection between this and the capitalist mode of production. In Wallerstein's work, ‘references to capitalist production do not point simply to market-oriented strategies for accumulating surplus value… the capitalist mode of production is a system in which groups pursue both political-military and profitable strategies, and the winners are those who effectively combine the two. Thus the state system, state building, and geopolitics are the political side of the capitalist mode of production’ (Chase-Dunn 1981, 25).

The longevity of the current world-system tends to be explained through the stability afforded by the series of hegemonic cycles, a profitable world production system and a relatively high measure of social cohesion in the core through the establishment of ‘liberal states offering suffrage, welfare, and a sense of racial/national superiority of its citizenry’ (Wallerstein Reference Wallerstein1996, 103). Indeed, for Wallerstein liberalism is a function of a state's hegemonic condition: in commercial policy it plays to their economic strength, and politically parliamentary institutions lessen the risk of violent, revolutionary change whilst curbing the arbitrariness of bureaucratic power (1983, 102–03). Wallerstein identifies three instances of hegemony in the modern world-system: the United Provinces (from roughly 1625 to 1672, the United Kingdom from 1815 to 1873, and the United States from 1945 to 1967) (1983, 102; see also Arrighi 2010, 6–7). The history of specific hegemonic orders is understood through deeper patterns of accumulation in the world economy. Whilst world-system writers employ a number of different cycles and sub-cycles, most famously perhaps the Kondratieff cycle rooted in measurement of production and prices (Goldstein Reference Goldstein1985), Arrighi develops the notion of ‘systemic cycles of accumulation’ to analyse the ‘formation, consolidation, and disintegration of the successive regimes through which the capitalist world economy has expanded from its late medieval sub-systemic embryo to its present global dimension’ (see Chase-Dunn and Grimes Reference Chase-Dunn and Grimes1995, 403–41; Wallerstein Reference Wallerstein1996, 98; Arrighi 2010, 6–10).

Hence, within each hegemonic ‘system-cycle’ the role of the leading state has been to form a hegemonic order that finds a way to combine capitalist economic and financial interests with the prevailing social and political forces of the age through the creation of a specific ‘regime’, or ‘social structure of accumulation’ (see Arrighi 2010, 10; see also Wolfson Reference Wolfson2003; Lippit Reference Lippit2005). This is an important concept to which the discussion will return below as it is the key to understanding how the tensions and violence endemic in the hierarchical world-system will manifest in a variety of forms – and not necessarily as inter-state violence – in different historical periods depending upon the social and political specifics of the contemporary accumulation regime.

The period of decline or disintegration for any particular cycle or regime of accumulation is marked first by a ‘signal crisis’ in which the prospect of accumulation through reinvestment in material expansion of the world economy begins to wane and capital turns instead to financial intermediaries and speculation. Whilst this switch can generate a temporary flourish of returns, it is actually serving notice of a deeper systemic crisis, which marks the ‘terminal’ crisis of the dominant regime of accumulation (2010, 220–21). In the present age, the crisis of the 1970s was regarded by Arrighi as marking the signal crisis of US hegemony and the Iraq War (2003) and the subsequent financial crises the terminal crisis. Such scenarios, however, should also be read in light of the long-term evolutionary dynamic in the world-system. This is apparent in the trend whereby each hegemonic order internalizes an increasing number of functions, manages an increasingly complex array of issues and problems, develops more extensive governance (including institutional) capacity and achieves greater global reach. It is in the interplay of these two dynamics that the world-system may be characterized as both cyclical and evolutionary.

Both Wallerstein and Arrighi are sceptical over whether, following US decline, a further hegemonic state will emerge (Wallerstein Reference Wallerstein1996, 102–03; Arrighi 2010, 374–78) and Arrighi identifies the trend of scale from city-state to nation-state to world state and raises the possibility that the ‘evolutionary process of simultaneous expansion and suppression of the modern interstate system might result in some form of world government due to the inability of even the most powerful state to provide sufficient governance in an increasingly chaotic world’ (2010, 76–77; see also Chase-Dunn and Inoue Reference Chase-Dunn and Inoue2012).

From this general discussion of world-system theory's ontological and conceptual bases, the discussion turns now to the two features at the edge of the democratic world that are potentially problematic for the Democratic Peace: the limited spread of democracy in the world-system and the incidence of conflict and violence between democracies and within the liberal zone of peace. With regard to the matter of democratization in the periphery, the discussion turns to whether structural factors and in particular relations of dependence limit the development of democracy.

Democratization in the periphery

Historically, the major trend has been for the expansion of democratic regimes, despite periods of setback and reversal such as in the 1930s. Whereas in 1974, 28% of the world's states were electoral democracies this had increased to 62% in 2008 (Merkel Reference Merkel2010, 22). Proponents of the Democratic Peace may take encouragement from this trend, especially given that democracy has spread to peripheral and semi-peripheral regions of the capitalist world economy, which appears to run against the view within world-system theory that the prospects for democracy in the peripheral zones were poor. Whilst historically, core states have had a variety of domestic political regimes (including e.g. absolutist monarchy), and there is no necessary theoretical connection between core status in the world economy and democracy at home, Korzeniewicz and Awbrey (1992) found that across three different time points, 1970, 1980, and 1990 virtually all core states enjoyed stable, highly democratic regimes whilst virtually all periphery states low democratic regimes. However, whilst the prevailing political regimes in the semi-periphery were in 1970 and 1980 either dictatorial or else short-lived democracies, ‘beginning in 1980, and clearly apparent by 1990… semiperipheral nations shifted in the direction of intermediate and/or stable democratic regimes’ (1992, 620–21). The liberal inference from this is obvious: that capitalism does not necessarily impede (and may in fact enhance) the spread of democracy beyond the core, thereby retaining the ultimate possibility of a universal liberal Democratic Peace evolving within the current structural formation of the international system.

Since the mid-1990s, however, the ‘third wave’ of democratization has stalled and the ‘voluntaristic’ approaches of the 1980s that stressed democratization as a matter of elite choice (see O'Donnell et al. Reference O'Donnell, Schmitter and Whitehead1986; Przeworski Reference Przeworski1986, Reference Przeworski1991) have lost ground to approaches that stress the significance of structural factors in influencing the prospects of democratization. More recently, voluntaristic approaches have argued that the emergence of democracy (but notably not the capacity to sustain it) is unconnected to levels of income (Przeworski and Limongi Reference Przeworski and Limongi1997; Przeworski et al. Reference Przeworski, Alvarez, Cheibub and Limongi2000; but see Boix and Stokes Reference Boix and Stokes2003; Epstein et al. Reference Epstein, Bates, Goldstone, Kristensen and O'Halloran2006). But even writers such as Francis Fukuyama who are usually associated with a liberal position have identified the exacerbation of economic and social inequalities along with weak institutions and poor US leadership as key factors in explaining setbacks to democracy (Reference Fukuyama2010; see also Carothers Reference Carothers2002; Diamond Reference Diamond2008). There should be no misunderstanding, however, for this is a complex area in which there is no general theoretical consensus and in which a number of factors would plausibly seem to shape the prospects for democracy and democratization (see, e.g. Haggard and Kaufman Reference Haggard and Kaufman2012). For an account of how participation in the capitalist world economy and relations of dependence affect political development, however, the work of the comparative political sociologists, Rueschemeyer et al. (Reference Rueschemeyer, Stephens and Stephens1992) and Huber et al. (Reference Huber, Rueschemeyer and Stephens1997) remains insightful. Their analysis complements the world-systems point that a peripheral position in the hierarchical world economy is likely to limit the expansion of democracy and, if it should emerge, to distort its political development.

For Samir Amin that ‘third world countries have almost never seen their political systems develop in the genuinely democratic manner (on the lines of the developed capitalist countries of the West)’ is due fundamentally to the incompatibility of democracy ‘with the demands of capitalist expansion’ (Reference Amin1993, 61). In the ‘third world’, the repressive nature and legitimacy deficit of authoritarian regimes will inevitably lead to domestic political challenges, but pressures for democratization are likely to lead to one of two diminished or incomplete forms of democracy. The first is the ‘populist’ response that addresses some aspects of the social problems that beset peripheral societies but which tends to be led by a charismatic leader who maintains a ‘top-down’ approach to political power such that the masses that support the regime are ‘not permitted to organize as an autonomous force in respect to the authorities’ (1993, 68). Such regimes may well pursue policies of industrialization, nationalization of sectors dominated by foreign capital, land reform, and reforms in the areas of health, education, and social rights. Their characteristic weakness, however, lies in their leadership structure and susceptibility to foreign destabilization and intervention as they challenge the economic and political interests of the core.

The second is the ‘petty democracy’ in which regimes recognize the principle of multi-party elections and grant a measure of free speech, but which fall short of addressing fundamental social problems or challenging relations of dependence. These regimes are unlikely to resolve the contradictions of democracy under capitalism for their responsiveness to the priorities of the core prevents pursuit of a social reformist agenda and exacerbates inequality and social problems, thereby maintaining the country in a state of instability and crisis. It is this second variant that corresponds to what Gills et al. (Reference Gills, Rocamora and Wilson1993) refer to as ‘low-intensity democracy’, which corresponds with Robinson's notion of ‘polyarchy’ discussed below. Beyond these forms of government states elsewhere in the periphery may be marked by the brutal ‘kleptocracies’ of the fourth world, communal ‘ethnic’ politics, or religious fundamentalism, often after earlier secular political movements and channels of dissent have been suppressed.

Rueschemeyer et al. (Reference Rueschemeyer, Stephens and Stephens1992) share with world-system theory an interest in the historical contextualization of the democratic turn, its structural dimension evident in the asymmetrically interdependent development of core and the periphery, and the propensity for similar processes to generate different outcomes depending upon whether these occur in the core or the periphery. For Rueschemeyer et al., a state's path of political development – and in particular the question of whether a state will develop along democratic lines – is a function of three clusters of power relations: (i) the balance of power among different classes and class coalitions; (ii) the structure, strength, and autonomy of the state apparatus and its interrelations with civil society; and (iii) the impact of transnational power relations on both the balance of class power and on the state–society relationship. One of the key conclusions when evaluating their extensive comparative analysis was that ‘transnational power relations – war, the structural effects of economic dependence, and economically and geo-politically conditioned interventions of foreign powers – profoundly affected chances for democratization’ (1992, 277). Whilst the degree to which transnational factors influenced the prospects for democratization varied across regions, ‘the key factor underpinning these differences was dependence in the world system, which had… generally unfavourable effects on the chances for democratic consolidation’ (1992, 278).

However, in a nuanced series of findings the authors differentiate between ‘geopolitical’ and ‘economic’ dependence which, in turn, has direct implications for appreciating variation in the space for democratic development according to the degree of geopolitical tension (see also Boix Reference Boix2011). To quote at some length,

The above passage repays close reading for whilst capturing the conventional world-system position on the influence of foreign capital on the class bases of states in the peripheral zones it also points to the transformative possibilities of the relationship between foreign capital and domestic political development, which have become more pronounced under conditions of globalization. Hence, whereas geopolitical dependence (such as economic and military aid or support for the military) ‘strengthened the state apparatus and allowed it to assume a high degree of autonomy from dominant as well as subordinate classes’ (1992, 278) economic dependence alone does not necessarily foreclose the possibility of democratic development in the longer run.

But it is likely to affect the form of democracy that emerges and the extent to which it is able to meet the needs of the citizenry. Huber et al. (Reference Huber, Rueschemeyer and Stephens1997) usefully differentiate between the ‘formal’, ‘participatory’, and ‘social’ dimensions of democracy. Many commentators, particularly in the core, have been content with the emergence of ‘formal democracy’, marked ideally by a political system that combines free and fair elections, universal suffrage, accountability of the state's administrative organs to the elected representatives, and effective guarantees for freedom of expression (1997, 323). However, in so far as formal democracy may be compatible with very limited citizen participation and massive economic and social inequalities it is unlikely to address the real needs of large parts of the citizenry, most notably in the peripheral zones, such that it amounts to little more than a democratic shell. ‘Participatory democracy’ by contrast entails ‘high levels of participation without systematic differences across social categories (for example, class ethnicity, gender)’ whilst ‘social democracy’ also includes increasing equality in social and economic outcomes’ (1997, 323–24; see also Weyland Reference Weyland2004).

Whilst the formal dimension of democracy provides a necessary framework for the check of arbitrary and corrupt government and protection of human rights – it is the participatory dimension that serves to guarantee and vitalize democracy through the involvement and mobilization of civil society and the social dimension that serves to counter-balance inequality, both of which are important for the consolidation and deepening of democracy. Yet for Huber et al. under conditions of neoliberalism international power structures ‘encourage formal democracy, while virtually blocking a deepening of democratic decision making and policies aimed at a reduction of social and economic inequality’ (1997, 338). If democracy fails to meet the needs and aspirations of the citizenry or perceived to operate to the benefit of a privileged group it will generate cynicism and become vulnerable either to challenges from alternative programmes or from a deeper societal malaise or fragmentation.

From a critical materialist perspective, the expansion of democracy – and increased support for the promotion of democracy internationally by core states – needs to be understood in relation to the neoliberal turn. For Robinson, ‘polyarchy’ as a model of democracy developed as an elite mode of rule in the late 19th and early 20th century as a way of neutralizing democracy's radical transformative potential in the hands of the rising labour movement (1996, 50). For similar reasons, democracy became favoured in US foreign policy circles during the late 1970s as the successor to authoritarianism in the third world. In the post-1945 period when Washington was seeking to consolidate and secure ‘a budding post-colonial international capitalism under US domination’ (Robinson Reference Robinson1996, 15) third world states tended to be primarily agricultural, commodity, or raw material producers with military or authoritarian government. Should local democratic pressures challenge this regime of accumulation the state had at its disposal a variety of internal and external resources with which to repress it. However, by the 1970s and 1980s globalization in the form of increased mobile, transnational capital in conjunction with developments in information and communication technology was fostering the development of a more complex and diverse economic structure in the periphery that was breaking down the utility and capacity of states to use repressive measures to resist pressures for democratization. These globalizing forces strengthened indigenous mass movements against repressive governments and at the same time led local industrialists and business leaders to re-evaluate their relationship to authoritarian regimes in light of their growing relationship with international capital (see Schwartzman Reference Schwartzman1998; Weyland Reference Weyland2004).

Whereas Washington elites saw that authoritarian regimes in Iran and Nicaragua were unable to avert revolutions that directly challenged US interests, in Spain and Portugal by contrast the successful turn from authoritarianism to polyarchic democracy offered a model that could potentially be applied elsewhere. For Robinson, US ‘democracy promotion’ marks the:

From a critical materialist perspective, then, structural factors have historically limited the development of democracy in the peripheral zones through complex interventions in their domestic social relations. With globalization and the neoliberal turn, the number of democratic regimes may have expanded but the shift from the ‘nation-state’ to the ‘neoliberal’ state marks the modification of democracy in the interests of capital and increasing governance by pro-capitalist elites. This in turn indicates a significant degree of structural variation in the type of democracy that prevails in the core and the periphery with that in the periphery in particular more akin to the product of hegemonic social relations than the idealized account of Democratic Peace theory as identified by Doyle above. That the Democratic Peace is through this lens a ‘hegemonic’ peace rather than the ‘just’ peace of liberal theory challenges both its legitimacy and its stability. That it is also prone to violence in areas where the core and periphery spill into the democratic world also calls into question its empirical validity.

Changing modalities of violence at the frontier of the Democratic Peace and the question of ‘exceptions’

From a world-system perspective, it is not only that the capitalist world economy limits the prospects for democratic political development, particularly in the periphery, but that as a residual source of tension it has generated conflict and violence between democratic states and between democratic states and democratic movements. The discussion below traces in outline some of the ways in which the ‘flow’ of this core-periphery violence has impacted upon the democratic realm. In some periods, core-periphery violence has been manifest in relations between democracies, and it is this form that has been most visible on the Democratic Peace radar, but in others it has been manifest in the suppression of constitutional or popular movements within imperial frameworks or as in the present period in the societal realm, reflecting shifts in the spatiality of core-periphery relations under conditions of globalization.

To begin, however, it is worth considering the category of ‘exceptions’ to the Democratic Peace, for as with the concepts of the ‘state’, ‘democracy’, and as will be shown ‘war’, how one understands such a key term shapes the interpretation and significance of the wider subject. Democratic Peace theorists were correct to regard certain nominal exceptions to the core proposition that liberal states or democracies rarely if ever go to war against one another as being of little significance. For example, that Finland was technically at war against the Allied Powers from 1941 to 1944 had little do with political tension between democratic states and much to do with Finland's desire to avoid being annexed by the Soviet Union (see Spiro Reference Spiro1994, 61–62; Elman Reference Elman1997). But if an exception or series of exceptions is rooted in clear political tensions between democracies, then this raises more serious concerns about the validity of the Democratic Peace. It is, then, the roots of conflict in political differences that is important for evaluating the question of exceptions to the Democratic Peace. In this case, it is differences arising from the political relationship of core-periphery that is of interest. Moreover, an emphasis upon the significance of the political character of prospective exceptions, as distinct from a formal or technical definition of war, brings the analysis into line with the understanding of war in the discipline more widely: war as the extension of politics by other means, as developed by Clausewitz.

This, in turn, may lead one to be circumspect about an over-reliance within the Democratic Peace upon formalistic accounts of what types of violence may and may not be regarded as exceptions to the Democratic Peace. In the Democratic Peace literature ‘war’ is commonly defined as organized institutional interstate violence resulting in over 1000 battle deaths (Russett Reference Russett1993, 12). However, ‘war’ so defined is but one modality of violence and as Barkawi and Laffey point out has never been the typical form of violence in relations between core and periphery (see Barkawi and Laffey Reference Barkawi and Laffey1999). Accordingly, one might reasonably be receptive to the inclusion of other politically significant categories of violence as comprising exceptions to, or at least be troubling for, the Democratic Peace.

The dyadic inter-state formulation of the Democratic Peace tends to overlook ‘extra-systemic’ violence, that is to say violence between core and periphery in the colonial era. Here one finds a range of proto-democratic movements, social revolutions, and partly democratic regimes that fought against liberal members of the core. In this period, the integration of local economies within the world economy and the penetration of capital was in many regions extensive and well established. There had already developed complex transnational relations of class, state, race, and capital as international influences upon local forces shaped the development of peripheral social relations in a variety of ways. Whilst the specific character of these relationships varied on a case-by-case basis, the ‘divide and rule’ principle of imperial government often contributed to local societal divisions, and ideologically imperialism went hand in hand with racist and sexist geo-cultures in order to reconcile or ‘make sense’ of the structural contradiction between universalism and hierarchy in this period (see Wallerstein Reference Wallerstein1996, 97–98).

The first of two illustrative conflicts is the British suppression of the ‘Egyptian Revolution’ in 1882. The revolution's slogan was ‘Egypt for the Egyptians’ and one of the key demands was a constitution that gave Egyptians (as distinct from Turks) greater political rights, not least over financial affairs and the appointment of higher-level military officers. However, as Reid notes, ‘Britain and France might be liberal democracies at home but they would not hear of constitutionalism in Egypt’ (Reference Reid1998, 221). Does this invalidate the claim that liberal states have not gone to war against one another? Of course not. But it is a clear indication of how liberal universal categories were sacrificed for the maintenance of an imperial international order, part of which was the ‘right’ of financial creditors to intervene and take control of a country's finances following a crisis or default. That Britain strangled Egyptian national self-determination in 1882 is, given the stage of Egyptian political development in that period, mutatis mutandis evidence of how the hierarchical relations of the world-system can override respect for democratic norms.

To turn to southern Africa in the same period, Gladstone had protested against the 1877 re-annexation of the South African Republic on the grounds that it put Britain ‘insanely’ in the ‘strange predicament of the free subjects of a Monarchy going to coerce the free subjects of a Republic’ (Robinson et al. Reference Robinson, Gallagher and Denny1981, 65). At this time, the Republic was under the Presidency of T.F. Burgers who is credited with having developed the state in a progressive manner. Following the First South African War with Britain in 1880–81 and the discovery of gold on the Rand in 1886 (and the subsequent social turmoil this presented for the republic), the state became increasingly autocratic under the administrations of President Kruger such that by the outbreak of the Second Boer War in 1899 the republic could not credibly be said to be ‘free’. However, the race question aside, its neighbour, the Orange Free State, was regarded by contemporaries as a ‘model republic’ and was much more integrated into the British imperial system. It had, however, become closer to its more autocratic republic out of fear that its own future independence was at risk following the Jameson Raid in 1895 and joined the South African Republic in war against Britain. The political development of both republics then, as well as their war involvement, is directly appreciable in terms of the wider imperial relations in which they were bound and in Britain's increasingly aggressive policies following the discovery of gold (see MacMillan Reference MacMillan1998).

To move forward, in the period from 1945 through to the 1980s two forms of violence that are problematic for the Democratic Peace are forcible covert actions and support for authoritarian regimes. The question of whether forcible covert actions by strong democracies against weak democracies can be counted as exceptions to the Democratic Peace thesis, and if so how significant they might be, was raised early in the research agenda but merits reconsideration given that from a critical materialist perspective these episodes are not part of a discrete ‘Cold War’ moment but specific manifestations of deeper historical and structural tensions (see Doyle Reference Doyle1983b; Forsythe Reference Forsythe1993; Russett Reference Russett1993; see also Barkawi Reference Barkawi2001; Rosato Reference Rosato2003). Of particular significance are six cases discussed by Forsythe (Reference Forsythe1993) and Russett (Reference Russett1993) in which the United States covertly used or threatened violence against elected governments in Iran (1953), Guatemala (1954), Indonesia (1957), Brazil (from 1961), Chile (1973), and Nicaragua (from 1984). The first of two points of contestation was whether the weaker democracy was sufficiently democratic to count as an exception to the Democratic Peace. The second point is one of interpretation as to whether the covert nature of these operations means they should be regarded principally as exceptions or vindications of the core thesis. It is this latter point that will be discussed here as a critical materialist reading furnishes a distinct interpretation.

From a liberal institutional-normative standpoint, Russett (Reference Russett1993, 123) attaches considerable significance to the covert, deniable character of these activities as indicative of the difficulties that administrations would have faced in gaining public support for them through regular institutional channels. Hence ‘in a very important sense, the U.S. democratic political system worked to limit intervention…. Normative/cultural and structural/institutional restraints were strong enough to forestall open military action, but not strong enough to prevent a secret operation or to stop it belatedly’ (1993, 124). However, the Central Intelligence Agency is an official, constitutionally regulated government agency and from a critical materialist perspective such operations do not signify an institutional loophole or evidence of incomplete democratic development, but rather the contradictions between the public discourse of democracy and the material-strategic nature of the modern capitalist state.

This contradiction between democracy and capitalism is a key factor in leading the liberal/democratic capitalist state to maintain the institutional apparatus to ensure the satisfaction of class interests when required. In this sense, such executive agencies are a structural requirement of the modern democratic state so that the state can achieve its objectives – and maintain the conditions of its own existence – in those circumstances when democratically elected legislatures either would not support such actions or would be seriously divided, or when the state itself would be publicly embarrassed. So, yes, as Russett rightly argues there is a normative dimension to this matter but what is lacking is an analysis of the underlying structural–material tensions that push the liberal state into potential or actual normative crises. Such operations were manifestations of the crisis that emerged when a defence of the capitalist world-system cannot be reconciled with democratically expressed anti-systemic pressures. That such actions form part of a specific ‘social structure of (capitalist) accumulation’ and serve a structural requirement of the democratic state is the basis for the inclusion, not exclusion, as exceptions to the Democratic Peace.

That democratic Powers have a long historical record of supporting authoritarian regimes in the peripheral zones is well established (Chomsky Reference Chomsky1993, 82–83; Schmitz Reference Schmitz1999). The specific point of engagement with the Democratic Peace programme is that democracies have frequently sponsored and maintained right-wing authoritarian regimes that have suppressed democratic opposition and violated human rights. Whilst many of these cases were during the Cold War, Schmitz has shown that US support for right-wing dictators emerged as a response to fears for order and stability following the First World War and the Russian revolution, not as a function of Cold War geopolitical rivalry but as a Republican counter-move to Woodrow Wilson's emphasis on self-determination. Yet even Wilson shifted his concern from that of autocratic governments to the containment of Bolshevism, which was a priority shared by subsequent Democratic as well as Republican administrations. Hence, following the First World War and the rise of nationalist and communist movements, ‘this emphasis on order came to permeate policy-making in Washington, and the United States found strong-arm rule, the maintenance of stability, anticommunism, and protection of investments sufficient reasons to support nondemocratic rulers’ (Schmitz Reference Schmitz1999, 6).

At the heart of such tensions between core and periphery were the popular and democratic aspirations of sectors of the population in the periphery poised against the attachment to order and property rights by core interests. Governments in the core, particularly liberal governments, have sometimes sought to address this tension through the search – generally to be disappointed – for ‘our kind of liberal’ in the periphery who would be politically ‘safe’ whilst pursuing a limited degree of social reform. However, the polarization of societies in the periphery has often meant that the kind of societal support base necessary for such a figure to succeed has been limited and any ensuing cooperation with more radical nationalist or socialist forces has been regarded with suspicion in the core. Accordingly, even figures such as Arbenz (Guatemala, overthrown in 1953), Bosch (Dominican Republic, overthrown in 1963), and Goulart (Brazil, overthrown in 1964) that were the closest one can find to reformist liberals in Latin America were regarded with suspicion due to their pursuit of an independent or non-aligned foreign policy, limited land reform and nationalizations, and refusal to vigorously suppress the Left. Following their overthrow, these leaders were replaced by US supported right-wing dictatorships.

But what of the standing of Democratic Peace in the current period in which democracy has spread and in which core Powers and International Organizations encourage the spread of democratic practices? It should be noted that there is some evidence that liberal norms have been consequential in limiting the capacity of large democracies to do harm to small democracies. For example, in 2002 following a domestic coup against the Venezuelan President, Hugo Chávez, Washington was embarrassed into standing by the ‘Collective Defence of Democracy’ regime grounded in a 1991 agreement of the Organization of American States. Chávez had become a thorn in Washington's side due to his rejection of neoliberal economics and wider ideological assault on the US's role in world affairs. Washington's initial failure to condemn the coup against the elected President was widely regarded as a clear expression of Washington's preference for the Venezuelan opposition, comprising the conservative business wing and sectors of the military and labour movement who favoured Washington's neoliberal programme. However, the Latin American members of the Organization of American States (OAS) roundly condemned the coup that subsequently imploded due to local developments on the ground. The episode led the United States subsequently to insist that any change of regime be undertaken through electoral means and with regional involvement through the OAS (Parish et al. Reference Parish, Peceny and Delacour2007, 223). This example of the Democratic Peace in operation is certainly significant, both empirically and theoretically, but a more critical reading challenges the relevance of this type of case in the present globalized age.

The process of neoliberal globalization since the 1970s has generated two particular trends that have implications for understanding Democratic Peace in this period. The first is the rise of the ‘neo-liberal’ state or ‘market democracy’ as discussed in the previous section. Such states, particularly in the more traditional peripheral areas are highly penetrated by a range of international institutions, organizations, and firms operating in conjunction with local actors and which exert significant leverage over matters of economic policy, political development, and governance. In such circumstances there is little need for core democracies to use direct violence as there are a range of indirect and non-military points of leverage that maintain an order favourable to the core.

However, the second trend is that the core-periphery frontier has shifted – or flowed – from the ‘classical’ political geography of dependency between first and third world states to a more chequered global spatiality, in which one can find elements of the core in the periphery and periphery in the core. Such a pattern is in-keeping with the capitalist logic of power identified above and is indicative of a shift in the balance of power between capital and state in favour of the former. From this shift in the geography or spatiality of the core-periphery, it follows that the political tensions and violence associated with this core-periphery interface will shift too. In this case, the shift has been away from the inter-state to the societal realm, which in recent years has become marked by greater levels of inequality and social fragmentation.

To measure the relationship between inequality, poverty, and violence, however, has not proved easy. Christopher Cramer argues that ‘inequality is hugely important to explaining civil conflict, but only insofar as the economic is considered inseparable from the social, political, cultural and historical’ (Reference Cramer2003, 409), taking issue with the methodology of large-n econometric studies that suggested little if any correlation. Cederman et al. (Reference Cederman, Weidmann and Gleditsch2011) find a statistical link between structural inequalities and civil war, which is applicable in cases both of economic and political inequalities between groups. For Sen (Reference Sen2008) as well as Cramer (Reference Cramer2003) and Cederman et al. (Reference Cederman, Weidmann and Gleditsch2011), the recognition of inequality as a causal factor in generating conflict is matched by the need to accept more complex or ‘messy’ causal scenarios.

According to Sen, ‘poverty and inequality are importantly linked with violence and lack of peace, but they have to be seen together with divisions in which other factors, such as nationality, culture, religion, community, language and literature, play their parts’ (2008, 12). Sen's comments that the political violence in France in 2005 was linked to economic and social marginalization and to factors of race (2008, 15) would seem equally applicable to the riots in London in 2011. But such explosive expressions of alienation do not of themselves convey the residual, simmering tensions of everyday life in deprived areas marked by high levels of social exclusion. The growing gulf between the life chances of a semi-permanent underclass, a precarious middle and the wealthy elite carries with it deep social and cultural implications. The rise of criminality, drugs, gangs, anti-social behaviour, and the exacerbation of racial, religious, and ethnic divisions with a simmering potential for conflict are characteristic features of the age (Wallerstein Reference Wallerstein1994, 15). This social violence is matched by the rise of more invasive and repressive state practices evident in, for example, the rise in surveillance powers and the shift from social-welfare strategies to the increasing criminalization and incarceration of the poor and particularly of black males within advanced post-industrial societies (Wacquant Reference Wacquant2001; Jinkings Reference Jinkings2011). This is not the form of inter-state war that would most obviously invalidate the liberal Democratic Peace, but by the same token it is surely not ‘peace’ and as an expression of deeper structural tends in the shifting spatiality of core-periphery tension is significant for the research programme, given that the Democratic Peace requires healthy democracies.

The search for ‘exceptions’ to the Democratic Peace, then, is faced with a moving target. The political analysis of world-systems theory speaks to the Democratic Peace through identifying the root of exceptions in the core-periphery dynamic, with specific forms explicable in terms of the tension between liberal ideational universalism and capitalist economic hierarchy. The specific modalities of violence through which these tensions are expressed are not, however, fixed and trans-historical, but vary in accordance with the prevailing political spatiality of the core-periphery tensions and respective structure of accumulation. In the colonial period, the contradiction between universalism and hierarchy was manifest in the repression of self-determination in the peripheral areas and the prevalence of racist, sexist geo-civilizational discourses such as to override normative universalism. In the post-1945 period, covert operations and support for authoritarian regimes were used to circumvent normative universalism. In the present neoliberal age, diminished notions of democracy have thinned normative universalism such as to render it anaemic in the face of the growing power of capital, manifest socially in terms of increasing inequality and fragmentation.

That these empirical phenomena are connected and relevant to the Democratic Peace, however, is only evident when one steps outside of the liberal paradigm and views the world from different ground. It is here that Barkawi and Laffey's point about not taking the concepts that enable the Democratic Peace for granted, or Wallerstein's about the need to analyse concepts in terms that set out their purpose, circumscribes their applicability, and specifies their respective theoretical framework help to identify wider patterns of violence that in their respective ways challenge the notion that a liberal (i.e. capitalist) Democratic Peace offers a viable model for a universal, let alone a ‘perpetual’ peace.