Fifteen years ago, Charles Kegley spoke of a neoidealist moment in international relations. This article examines how the number of armed conflicts has declined in the decade and a half since Kegley’s presidential address and shows that the severity of war has been declining over a period of over six decades. The number of countries participating in war has increased, but this is in large measure due to coalition-building in several recent wars. Overall, there is a clear decline of war. It seems plausible to attribute this to an increase in the three factors identified by liberal peace theorists: democracy, trade, and international organization. Four alternative interpretations are examined: the temporary peace, the hegemonic peace, the unsustainable peace, and the capitalist peace. The article concludes that the latter, while running close to the liberal peace interpretation, also presents the greatest challenge to it. Indeed, we seem to be living in a commercial liberal period rather than a world of neoidealism.Footnote 1

1 Introduction

Fifteen years ago, exactly to the day, one of my predecessors as President of the ISA, Charles Kegley, alerted us to what he perceived to be a liberal moment in international relations. Or so I thought until I looked up the published version in International Studies Quarterly (Kegley 1993).Footnote 2 In fact, Kegley used the term ‘neoidealist’ rather than liberal, and there was a question mark in his title.Footnote 3 I cannot tell this particular audience that this proves the value of not destroying a good story by checking your sources. But I will stick with my own version for the time being, and return to the question of idealism at the end.

Whether under the heading of idealism or liberalism, it was quite visionary 15 years ago to talk about an emerging international order that might give us a more humane and peaceful world. The Cold War had just ended. But rather than producing peace in Europe, this had reopened old wounds in the Balkans and in the Caucasus. The long-standing armed conflicts in Northern Ireland, Kurdistan, and the Basque territory remained unsolved. In addition, Romania, Moldova, Turkey, Nagorno-Karabakh, Azerbaijan, Chechnya, Moscow, and three places in Georgia saw armed conflict over just a few years at the end of the Cold War.

Three armed conflicts broke out in former Yugoslavia, including the war in Bosnia, which was the bloodiest in Europe since the Greek civil war in the late 1940s. Srebrenica, with the murder of over 7,000 Bosnian men, was still two years down the road (Brunborg et al. 2003). Realists were warning that this was merely the beginning. We were going ‘back to the future’ or perhaps more appropriately, forward to the past. French-German rivalry would once again play up. The Germans were advised to acquire nuclear weapons to deter the French force de frappe (Mearsheimer 1990). Several contentious issues had arisen between Russia and newly independent Ukraine—the territory of Crimea, the Russian diaspora, the fate of the Soviet Navy, and last but not least, the nuclear arsenal. The realist advice to Ukraine was to hang on to some of the Soviet nuclear weapons in order to deter the Russians (Mearsheimer 1993). On a smaller scale, trouble was brewing between Hungary and several of its neighbors, which host a Hungarian diaspora some 25 % of the population of Hungary itself—in particular with Romania, which under Ceauşescu had actively persecuted them. A minor ‘water war’ was foreseen Hungary for and Slovakia over the Gabcikovo-Nagymaros dam project on the Danube. Figure 8.1 displays the actual and potential hotspots.

Fig. 8.1
figure 1

Actual and predicted conflicts in Europe at the end of the Cold War. Sources Actual armed conflicts 1989–94 (*) are from the UCDP/PRIO conflict data, cf. Gleditsch et al. (2002) and www.prio.no/cscw/armedconflict. The predicted conflicts (x) are based on my own reading of various media sources at the time. The figure was created by Siri Rustad.

More generally, the number of ongoing state-based armed conflictsFootnote 4 had reached a peak in the two years prior to Kegley’s address, with 52 armed conflicts active in 38 countries (Gleditsch et al. 2002; Harbom et al. 2008). The number of new conflicts also peaked in the early 1990s.Footnote 5 The world did not look like a peaceful place.

2 Less Conflict

No sooner had Charles Kegley announced a possible neoidealist moment than the number of conflicts started to decline, eventually settling to around 30, a level lower than at any time since the mid-1970s (Fig. 8.2).Footnote 6 Many conflicts hover just around the threshold of 25 battle deaths per year, so the list of conflicts is not stable from one year to the next, but the number remains about the same. However, after World War II there was a major expansion in the number of independent countries—‘the interstate system’ in the words of another presidential predecessor, J. David Singer. Internal conflicts in non-independent territories are generally ignored in compilations of armed conflict, so we can easily get an inflated picture of the rise of armed conflict during the Cold War. If we divide the number of armed conflicts by the number of independent countries, we get a much less steep increase up to 1991 and a steeper decline since then, to a level not observed since the early 1960s. The number of wars, in the sense of armed conflicts with a minimum of 1,000 battle deaths in a given year, has declined to a quarter of the peak level. There are very few very violent conflicts, but some long-standing low-violence insurgencies persist.Footnote 7 In the words of John Mueller (2004), it is mostly ‘remnants of war’ that are left—opportunistic predation by criminals, bandits, and thugs. The year 2007 was also the fourth year in a row with no recorded interstate conflicts.

Fig. 8.2
figure 2

The development of conflict, 1946–2006. Sources Number of conflicts and the number of countries participating in conflict based on the UCDP⁄PRIO data, see Gleditsch et al. (2002) and www.prio.no/cscw/armedconflict. Number of battle deaths (so far updated only until 2005), see Lacina/Gleditsch (2005) and www.prio.no/cscw/cross/battledeaths. The three graphs are set to 100 % in 1975, corresponding to 29 conflicts, 37 countries in conflict, and 135,653 battle deaths. Figure created by Siri Rustad.

Another encouraging sign is that the number of entirely new conflicts has declined even more drastically, to the point where no new conflicts started in 2005 or 2006 (Elbadawi et al. 2008, Fig. 8.2).Footnote 8 It is possible, of course, that the remaining conflicts, many of which are decades old, are harder to end than the others. However, as expressed so well in the title of a book by Fred Iklé ([1971] 1991), Every War Must End. This may take time, but as long as a war that ends is not replaced by a new war,Footnote 9 we will continue to see a decline in belligerence.

Several alternative measures of conflict show essentially the same picture. For instance, there is also a decline in the share of countries affected by war in their territory. However, one measure appears at first glance to give a different picture. This is the number of countries participating in armed conflict, also plotted in Fig. 8.2. A recent survey of armed conflict (Hewitt et al. 2007) makes a point of this.Footnote 10 The main reason why this indicator has not declined and has actually reached a historical peak is the coalition-building in the Gulf War of 1991, the Kosovo War of 1999, and the more recent wars in Afghanistan and Iraq. These four conflicts have from 20 to 36 participants. Compared to the Korean War with 20 participants, and the Vietnam War with just nine participants, these seem like very large wars. At 36 participants, the Iraq War is comparable to the two World Wars using this measure of size. Any other measure, of course, shows otherwise. Many of the participants in the recent wars, such as Iceland or Tonga,Footnote 11 have probably joined more as a matter of political solidarity than because they can make a real military contribution. Some of them have suffered no casualties.

Two of these wars, the Gulf War and the Afghanistan War after 2001, were sanctioned by the United Nations. The other two were not, but the United States still went to great lengths to acquire institutional backing from NATO in the case of Kosovo and from a more informal ‘coalition of the willing’ in the Iraq War. It would certainly be a stretch of imagination to characterize the invasions of Afghanistan in 2001 and Iraq in 2003 as ‘peacekeeping.’ However, for most of the countries now participating in military operations in these countries, the peacekeeping or peacemaking perspective is probably dominant. In that sense the rise in the number of countries nominally ‘at war’ is a questionable indicator of a resurgence of war. It is more consistent with the concurrent increase in international peacekeeping that we have seen in the same period (Fig. 8.3), an increase almost wholly due to a rise in peacekeeping activities in internal conflicts. The number of personnel participating in such missions has also risen remarkably. But it is hardly reasonable to interpret this as increasing global warfare.

Fig. 8.3
figure 3

The growth of peacekeeping, 1948–2008. Sources Figure created by Siri Rustad based mainly on data from Heldt/Wallensteen (2007), supplemented by Han Dorussen with data from the United Nations. The figures are compiled on a monthly basis, and we have used the January figures for each year

A more pertinent indicator of warfare is the severity of war, or the number of people killed in battle. The trend for the 20th century is completely dominated by the two world wars (Lacina et al. 2006: Fig. 8.2). The severity of war clearly peaked in the first half of the 20th century. After World War II the battle deaths continue to be heavily influenced by individual wars, but the peaks are declining (Fig. 8.2). The first peak is the Chinese civil war closely followed by the Korean War; the second is mostly due to the Vietnam War; the third represents the added effects of the Iran-Iraq War and the Soviet Afghanistan War; and in the fourth, we find the internationalized civil wars of the Democratic Republic of Congo. Figures for war casualties are often highly disputed. The Iraq War is a case in point. An extensive WHO survey (IFHS 2008) puts the number killed at 150,000 for the period March 2003-June 2006.Footnote 12 Iraq Body Count (www.iraqbodycount.org), which records published deaths, tallied 48,000 in the same period. The WHO figure probably includes some deaths from types of violence that have not been included in the numbers for earlier wars.Footnote 13 But even if we use the WHO figure as battle deaths, they do not reverse the long-term trend toward a lower lethality of war. Currently (Spring 2008), according to the Iraq Body Count’s tally of recorded civilian deaths, violence in Iraq seems to be declining. Nevertheless, it remains the most violent enduring war anywhere. Perhaps the Iraq conflict will settle into something like a much more violent version of Northern Ireland. In any case, it is certainly a great tragedy for those involved, as are the conflicts in Darfur, in Colombia, and in the other 30 plus locations. But none of them represents a reversal of the waning of war after the end of the Cold War. In the longer-term perspective, it still seems more appropriate to talk of a world climbing down from a peak of armed violence in the middle of the 20th century.Footnote 14

Statistics of state-based armed conflicts and their consequences in terms of battle-deaths do not tell the full story of human violence. I will leave out violent crime,Footnote 15 but deal briefly with three other missing elements. The first concerns indirect deaths in war. ‘Civil wars kill and maim people—long after the shooting stops,’ as another former President of ISA, Bruce Russett, and coauthors said in a seminal article (Ghobarah et al. 2004). Such effects include: revenge killings in the wake of the war; destruction of physical and human capital; slow economic growth; weaker social norms and political chaos; weapons proliferation and crime; increased flows of refugees and internally displaced persons; and environmental destruction including the littering of the landscape with landmines and cluster weapons. A major World Bank study (Collier et al. 2003) aptly characterized civil war as ‘development in reverse.’ And such consequences do not only affect the countries at war, but frequently also their neighbors.

But such indirect effects of war are not new to our age. The influenza epidemic that followed in the wake of World War I claimed some 40 million lives, more than the war itself (Riley 2001). The war contributed to the spread and lethality of the disease by increasing geographical mobility and by lessening resistance to illness, but we cannot say with any accuracy how many would have died if the war had not contributed. The same problem applies to more recent wars such as those in the DRC and in Sudan. It is possible that indirect effects of war are now greater relative to battle deaths because most armed conflicts take place in poor societies with weak health facilities. But we have no reliable time-series data to back up such a conjecture. The efforts of the international community to ban certain types of weapons, most recently cluster munitions,Footnote 16 may also reduce some of these indirect effects in the future.

A second omission is non-state conflict, organized groups fighting each other but without the state being a direct party to the conflict. The Uppsala Conflict Data Program records roughly the same number as for state-based conflicts, but they generally involve far fewer fatalities.Footnote 17 We do not have comparative data over a long period to establish clear trends, but in sub-Saharan Africa (where most of them occur) both their number and their lethality dropped substantially over the period 2002–06 (Mack 2008: 36). Given their low severity, they do not appear to offer a serious challenge to the idea of a waning of war.

The third and most serious omission concerns one-sided conflicts, that is, genocides, politicides, and, more generally, fatal attacks on unorganized people by governments and other organized groups. Many of these, such as the Holocaust or the liquidation of the Kulaks, are extremely severe and rank with the largest of wars. Most studies focus on individual conflicts or on a short time-range.Footnote 18 The best long-range data set is probably the one generated by Rudolph Rummel (1994, 1997) on what he calls democide, defined as ‘the murder of any person or people by a government.’Footnote 19 It includes ‘death by virtue of an intentionally or knowingly reckless … disregard for life …’ Examples include deadly concentration camps, medical experiments on humans, and famines or epidemics where the authorities withhold aid or even act in a way to make the situation worse. Clearly, Rummel includes deaths that in studies of war would be classified as indirect deaths. The democide data therefore cannot be compared directly with battle deaths in war. But assuming the criteria are reasonably consistent over time, his data show the same inverted U-shaped curve as for battle deaths, peaking in the middle of the 20th century (Rummel 1997: Table 23.1). Given the recent critical examination of China’s ‘Great Leap Forward’ and other disasters under Maoism, it is possible that the peak should be higher, and located later. But the downward trend in recent decades would still hold. The genocides in Rwanda in 1994 and more recently in Darfur, tragic as they are, also do not change the basic shape of the curve.

Tracking all sources of deadly violence is a tall order. Rather than analyzing them one by one or trying to add the number of deaths from the different sources, we may look to life expectancy as a good overall measure of lives not lost. Life expectancy has increased steadily over the last 200 years and is expected to continue to increase for the next half-century (Goklany 2007; Riley 2001; UN 2007). This applies to the world as a whole and to all regions but two: The exceptions are sub-Saharan Africa (mainly because of the AIDS epidemic), and Eastern Europe (because of Russian lifestyle diseases and economic collapse), though the UN projects increasing life expectancies for these regions as well.

The world average for life expectancy has increased from 26 years in the early 19th century to over 65 today (Maddison 2001). In this way, the world has gained many more years of life than lost through war and genocide. And our lives are not only longer, but also healthier (Goklany 2007: 40).

Historians and anthropologists, not to mention archeologists, will be dissatisfied with any reference to data from just the 20th century as ‘long-term.’ Although we have a large literature on earlier wars, it is difficult to find hard data that would enable us to do systematic comparisons over time. The Correlates of War Project has taken data on wars and civil wars back to the Congress of Vienna in 1815, but even simple comparisons of the number of wars and casualties become problematic. The first four decades after the Congress of Vienna had very few battle deaths according to the Correlates of War Project, but how meaningful is that information when the interstate system in 1816 consisted of just 23 states (Small/Singer 1982: 47–50)? My immediate predecessor as President of ISA, Jack Levy, has informed us that the number of great-power wars has declined in the last 500 years (Levy et al. 2001: 20), indicating a longer trend in the decline of war.Footnote 20 The mass murder of civilians is not a new phenomenon either. Genghis Khan, who is the common ancestor of 0.5 % of all males in the world, according to a recent genetic study (Zerjal et al. 2003), is widely credited with killing over a million people in the Muslim kingdom of Khwarezm in 1220—21.Footnote 21

Judging the long-term development of massacres and wars becomes even more difficult when we move to the pre-historical period. Several anthropologists have argued, in my view convincingly, that the idea of the ‘peaceful savage’ must be definitively discarded (Gat 2006; Keeley 1996; LeBlanc/Register 2003). Of course, largely peaceful communities can also be found (Fry 2006). War is not intrinsic to human nature, but neither is peace. The decline of violence may be much more of a long-term phenomenon than our statistical data indicate. I am nevertheless inclined to think that the peak of armed violence in the middle of the 20th century is real. We have lived through a particularly lethal combination of the old perception of war as a useful instrument of policy with the modern technological capability to wage war effectively. Our technological skills have continued to improve, so we could kill each other many times over if we applied the full range of human ingenuity to that task. A single direct nuclear exchange between the two superpowers would have changed the picture dramatically and created a more recent and higher peak of war severity. If we do not kill each other at such a rate, it is because our institutions and attitudes have changed. I conclude that we do seem to be moving toward a more peaceful world. But is it a liberal peace?

3 The Liberal Peace

When Charles Kegley addressed the ISA 15 years ago, the slogan of a liberal peace had not yet been coined, although the key liberal ideas about international relations had reached a venerable age. Karl Deutsch, who surprisingly is not a former President of ISA but whose name adorns one of our awards, had written about international security communities, held together without the use or even the threat of force (Deutsch et al. 1957). By 1993, a systematic empirical research program on the democratic peace had been under way for over a decade, initiated by Rummel (1983) and Doyle (1983) and later followed up by Maoz/Russett (1993) in particular.Footnote 22 The first systematic empirical case for a liberal peace can be dated precisely to 1996 (Oneal et al. 1996) and has been developed at great length in a series of articles and in a book by Russett/Oneal (2001). In fact, the Russett-Oneal project has become one of the most sustained and wide-ranging empirical research efforts to be conducted in any area of international relations. The two have always been very generous in sharing their data, even before IR journals started making this a requirement for publication (Gleditsch 2003). This has led to a number of new challenges to their work. Over the years, then, we have seen improvements in their model, their empirical measures, and their analyses. But they consistently find support for the liberal tripod: democracy, economic integration, and international organization. The literature disagrees on the relative importance of the three factors; some find that given democracy, trade has little importance (Beck et al. 1998), whereas others conclude the opposite (Polachek 1997). But the joint importance of the three factors seems to be well established.Footnote 23

Although the Russett-Oneal work in this area has focused on interstate war, in the tradition of Norman Angell (1910), more recent studies have also established the importance of democracy and trade for civil war (Blanton/Apodaca 2007; Bussmann/Schneider 2007; Hegre et al. 2003).Footnote 24 Democracies rarely experience large-scale civil warsFootnote 25; although, some have suffered long and drawn-out secession conflicts at a lower level, particularly where they are fueled by the promise of riches from oil and other raw materials or by religious and ethnic polarization. Much of the popular literature about globalization has emphasized its divisive nature, creating inequalities, and distributional conflict. But most empirical studies show that globalization in fact reduces armed conflict, if not directly then indirectly through its wealth-generating effects. Interestingly, even Katherine Barbieri, one of the first to challenge Oneal and Russett on the interstate liberal peace (Barbieri 1996), has found globalization to reduce the risk of civil war (Barbieri/Reuveny 2005). Since the overwhelming number of conflicts today are internal conflicts, this bodes well for the future of the liberal peace as long as the three liberal factors remain on the rise.

For interstate conflict we cannot as easily generalize from the dyadic to the systemic level. Virtually all the work on the interstate liberal peace is at the dyadic level. Russett and Oneal and others have shown that two countries that share a democratic system trade more, and have more ties through international organizations are less likely to fight. But this does not necessarily mean that a world of more democracies, higher trade, and a proliferation of international organizations will produce world peace. In theory, countries could refrain from fighting fellow democracies and their most important trading partners while still maintaining a high level of conflict involvement. In that case the liberal peace would imply a displacement rather than a reduction of warfare. Studies of the systemic effects of democracy provide an ambiguous answer. Of course, if democracies never fight one another, a world with 100 % democracies should have eliminated war. But at lower levels of global democracy the relationship is not so obvious (Gleditsch/Hegre 1997; Mitchell et al. 1999). In a world, or even a region, with no democracy, the emergence of a single democracy might in fact lead to more conflict. The process of democratization itself can also lead to instability and conflict (Mansfield/Snyder 1995; Ward et al. 1998). On the other hand, Mitchell (2002) finds that an increasing proportion of democratic states in the international system promotes the use of democratic norms even by nondemocratic states.

In a rare study of trade relations at the systemic level, Maoz (2006) found that trade interdependence has a consistent dampening effect on the amount of systemic conflict.Footnote 26 Lacina et al. (2005) found a statistical relationship between the three liberal factors and the decline in the severity of war, including fatalities in interstate as well as civil wars.Footnote 27 However, studies at the system level have few control variables. Generally, the liberal interstate peace is the least well established precisely at the level where it is most important. This has not prevented a range of policy makers from Ronald Reagan to Kofi Annan from embracing the liberal peace, and particularly the democratic peace component. They are probably correct, but it would be reassuring to have more studies at the system level.

Figure 8.4 graphs the development of the three liberal factors over time, normalized around the level in 1973. IGO membership has been increasing almost linearly since the end of World War II. Trade as a share of GDP has also increased most of the time but has exploded since the early 1970s. Finally, democracy has gone through its ‘waves’ (Huntington 1991) and is now at a level never exceeded, whether measured as the fraction of states under democratic rule or the percentage of world population living in a democracy (Gleditsch/Ward 2006). The rise of the three liberal factors is consistent with the recent decline in the number of wars and the longer decline in the severity of war. But these five curves do not match each other in any simple or convincing way for the entire period since 1945. In the 15 years since Kegley’s presidential address, the three liberal factors have gone up and conflict has gone down. But they did not turn around at the same time. The liberal factors were also increasing in the 15 years before his address, while the number of conflicts was rising. The growth of the liberal factors, with a partial exception for democracy, is more consistent with the long-term decline in the lethality of war.

Fig. 8.4
figure 4

The growth of the liberal factors, 1946–2004. Source For each of the four indicators, 1973 = 100 %. Sources For data on democracy: Marshall/Jaggers (2003). Polity IV Project, www.systemicpeace.org/, for trade⁄GDP: Gleditsch (2002), and for IGOs: Pevehouse et al. (2004). Original graph created by Bethany Lacina for Lacina et al. (2005), updated by Kristian Skrede Gleditsch

Despite the widespread acceptance of the idea that the decline of war is related to a liberal peace, there are also a number of alternative interpretations, some of them considerably less optimistic. I will examine four of them here.Footnote 28

4 Four Challenges to the Liberal Peace

4.1 The Realist Challenge: The Temporary Peace

The major challenge to the liberal peace still comes from the realists. Indeed, Kegley (1993: 143) ended his talk by advocating ‘development of a principled realism emphasizing liberal ideals.’ For realists, the international system remains anarchic, and its ups and down are determined by the state struggle for survival. Hence the most important variables are the economic and military strength of major powers and the patterns of alliances. Periods of greater or lesser peace will be explained in terms of the variation of these factors. Realists can argue that the post-World War II period was more peaceful than the previous decades because of its bipolar nature, making a direct confrontation between two blocs armed with nuclear weapons too dangerous. Instead, rivalries were channeled into proxy wars on behalf of the two major blocs. Realism is also sufficiently flexible to account for the decline in global warfare after the Cold War as a result of an even more stable unipolar order, where there is no real challenge to the hegemon. In the words of Fukuyama (1989: 4), democracy and the market economy were ‘the final form of government’ and thus defined the ‘end of history.’ But of course the realist factors cannot account in any direct way the ups and downs in the number of conflicts since World War II. Even leaving that aside, the more interesting question is how a real challenge could emerge to the seemingly hegemonic liberal system. It is simple enough to predict the slow relative decline of the United States as the one and only hegemon. Demographic factors and the phenomenal economic growth of China and India dictate that at some point the US economy will be overtaken and other countries will be able to purchase a more powerful military if they so desire. But predicting the slow relative decline of the United States is not the same as predicting the fall of the liberal peace. All the major challengers seem to have embraced the market economy. In three decades, China has moved from being a warfare state, internally as well as externally, to being a trading state, in the words of Richard Rosecrance (1986), another former ISA President. Politically, it has remained a one-party state, with frequent violations of human rights, but without the excesses of the Great Leap Forward or the Cultural Revolution. It seems to be experimenting with competitive elections at the local level and the wisdom of the party leaders is regularly questioned. Corruption is rampant, but corrupt leaders are also regularly being exposed (Thornton 2008). The time when China will be a reliable partner in a democratic peace with its neighbors seems distant, but the economic incentives for maintaining peace appear very robust. Its undemocratic leaders benefit as much from the present trading boom as does the general public, if not more.

Where else can we find a challenge to the hegemony of the market economy? In the remnants of communism in North Korea or Cuba? In the gerontocracy of Zimbabwe? Or among former military coup-makers like Hugo Chavez who can afford to play democrats as long as they can use abundant oil income to boost their position? Such regimes and rulers may have considerable nuisance value to the hegemonic states, but can hardly present a coherent global challenge to the hegemonic system in the same way that communism and fascism did from the 1920s onwards.

The major challenge seems to be found not in traditional economic or military power but in spiritual and cultural power, backed by historical memories. In that sense, it is not a head of state but an opposition leader, Osama bin Laden, who is the main challenger to the international order. Ethno-religious conflict was one of the leading candidates to fill the gap left by the end of the Cold War (Kaplan 1994). Gurr (1994), in the ISA presidential address following Charles Kegley’s, referred to a surge in ethnopolitical conflict after the end of the Cold War. However, he did not think there was a strong global force for the further proliferation of such conflicts, and a few years later he proclaimed that ethnic warfare was on the wane (Gurr 2000). Mueller (2000) dismissed the increased concern with ethnic conflict as ‘banal.’ The general ‘clash of civilizations’ predicted by Huntington (1993), with the civilizational fault lines largely determined by religion, has hardly been a dominant factor of world politics (Russett et al. 2000) and certainly has not reversed the waning of war. Nevertheless, a number of the major ongoing conflicts, such as the ones in Afghanistan, Iraq, and Sudan, have an important religious element. Indeed, Fig. 8.5 shows that since 1990 an increasing share of the world’s armed conflicts have involved Muslim countries, Islamic opposition movements, or both. But this is not due to an absolute increase in what we might call ‘Islamic conflicts’; their number remains relatively constant. It is the decline of other types of conflict that creates a relative rise of Islamic conflicts. In other words, in the general trend toward more peace, Muslim countries and Islamic opposition groups seem to be lagging behind, just as Muslim countries in general (and Arab countries in particular) are lagging behind in the rights of women and human rights more generally,Footnote 29 in democracy, in the eradication of illiteracy, and in the second demographic transition (UNDP 2005).

Fig. 8.5
figure 5

The number and share of armed conflicts involving Muslim countries or Islamic opposition movements or both, 1960–2006. Source Figure created by Halvard Buhaug on the basis of the UCDP/PRIO conflict data and his own coding of Islamic conflicts

This is not a clash of civilizations. Most of these ‘Islamic conflicts’ pit Islamic opposition movements against the governments of Muslim countries. Although the Iraq War of 2003 was an invasion of a Muslim country by a coalition composed largely of Christian nations, the government of Iraq was a secular, not a religious dictatorship. The Gulf War started because one Muslim country invaded another, as did the Iran-Iraq War. The specter of a mutual crusade or Jihad between Christians and Muslims certainly exists in the minds of many people. We may even fear that it becomes a self-fulfilling prophecy, but it remains a very incomplete description of today’s global pattern of conflict. Moreover, there is no evidence that religious conflicts are bloodier than other armed conflicts (Nordås 2007).

Perhaps the greatest realist challenge is what Russett (2005) has called ‘bushwacking the democratic peace’. The peace between liberal states has tempted major liberal powers to attempt to help the process along by force. Democracies tend to win the wars in which they participate, and when autocracies lose wars there’s a high probability of regime change, which frequently will go in the direction of greater democracy. In a sense, liberal and realist motivations become one and the same. If the West could democratize the Middle East, the liberal audit would be favorable, but so would a traditional security calculation of how to reduce the fear of spreading conflict and the threat to local allies. In this regard, liberals have regarded with some trepidation the lip service paid to the democratic peace by Margaret Thatcher and a series of US Presidents. With the invasion of Iraq in 2003, the worst fears of many liberals seemed to have come to pass. Early empirical work on forced democratization did actually find that military intervention by democracies resulted in some democratization (Hermann/Kegley 1998; Meernik 1996; Peceny 1999). Research conducted after the Iraq invasion, including that of another former ISA President (Bueno de Mesquita/Downs 2006; see also Pickering/Peceny 2006), has been more skeptical. Forced democratization usually fails to bring the new democracies to a very high level; rather, they tend to end up in the semi-democratic category where political instability and internal conflict is higher than in autocracies (Gleditsch et al. 2007). In the Iraq case, there is an additional reason for skepticism: Even if democratization had been successful, the new democracy would have been surrounded by nondemocracies, a mix for which democratic peace theory does not predict a peaceful future. The only way to overcome that problem would be to extend the policy of forced democratization to Iran, Syria, and others, further strengthening the alliance between liberalism and realism.

One reason why democracies are good at winning wars when they join them is that they are usually more successful than the other side at building alliances, as noted earlier. Even the Iraq War, opposed by many US allies, gathered a ‘coalition of the willing’ of no less than 36 countries. For liberals, the dilemma is that such coalitions usually include many illiberal states. This is a continuous theme from the Western wartime alliance with Stalin, who killed more people than Hitler, to the US alliance with Saudi Arabia, whose rejection of liberal values is just as firm as those of the enemies of the United States in the region. As Franklin D. Roosevelt reportedly remarked about Rafael Trujillo, dictator of the Dominican Republic, ‘He may be an SOB, but he’s our SOB.’ (Paterson et al. 2005: 157).Footnote 30

4.2 The Radical Challenge: The Hegemonic Peace

A second challenge to the liberal peace is the radical interpretation. It agrees with the realist view in interpreting the current decline of conflict as mainly a result of an overwhelming hegemony on the part of the West in general and the United States in particular. Thus, the current peace is a hegemonic peace or an imperial peace (Barkawi/Laffey 1999). Unlike realism, the radical view focuses on social and economic inequalities within and between societies. In this view the current peace is the peace of the successfully run plantation where the slaves go about their business without questioning their circumstances. Dependency theory has depicted economic relationships between the center and the periphery as exploitative, where elites on both sides ally against the underdogs (Bornschier/Chase-Dunn 1985; Galtung 1971). Sooner or later, however, the downtrodden are going to rise against domestic elites and international hegemons. In the 1990s, violent street demonstrations in Seattle, Prague, and elsewhere signaled the solidarity of anti-globalization forces with the global underclass. The Marxist states were largely gone, but Marxist analyses of world politics were not.

Much of the anti-globalization literature builds on the premise that in the neoliberal world, inequalities have been rising within, as well as between, countries. Indeed, Charles Kegley (1993: 140) noted in passing ‘the widening gap between the world’s rich and poor.’ In fact, on a global basis, individual economic inequality has undergone a massive decrease, thanks to the phenomenal economic growth of poor countries like China, India, Vietnam, and many others (Firebaugh 2003; Goklany 2007; Neumayer 2003). During this process inequality within these same countries has increased, not through impoverishment of the masses, but because while wealth is created many people remain left behind in poverty. Thus, we have an unequal peace.Footnote 31

At the global level, Paul Collier’s (2007) recent book The Bottom Billion argues that the world is making major progress in promoting development, but that one sixth or so of mankind are left out of this process. Around 60 countries suffer not only from low GDP per capita but also low or negative growth. These countries tend to be caught in one or several development traps: armed conflict, the resource curse, being landlocked with bad neighbors, bad governance, or being too small. Unfortunately, Collier will not reveal his list of countries,Footnote 32 but by applying his criteria, we find something like the map in Fig. 8.6. By superimposing the ongoing armed conflicts for 2006, we see a certain overlap, but it is by no means perfect. Perhaps if the bottom billion notion had taken more account of disparities within countries, there might have been a closer fit.

Fig. 8.6
figure 6

The Bottom Billion countries and armed conflicts in 2006. Source The Bottom Billion data were collected by Åshild Falch. The conflict map was created by Halvard Buhaug and Siri Rustad

The literature in political science and economics is divided on the effects of such inequalities. The relative deprivation tradition has pointed to inequality as a cause of conflict (Gurr 1970); although, some have argued that the evidence was inconclusive (Lichbach 1989). Cross-national studies of overall income inequality in a society (so-called vertical inequality) and civil war tend not to find any significant relationship (Collier/Hoeffler 2004; Hegre et al. 2003). But some recent work points to the importance of horizontal inequality in promoting conflict–that is, socioeconomic or political inequalities between groups, such as ethnic or religious groups in generating internal conflict (Østby 2008; Stewart 2002). With increasing inequalities in many countries, this may well be an increasing source of internal conflict. However, it is hardly the stuff of which major wars are made.

4.3 The Environmental Challenge: The Unsustainable Peace

Many environmentalists take a dim view of the future and man’s exploitation of the natural resource base. This is an old story that goes back to the Malthusian problem of matching food production to population growth. Malthus (1798) thought that this inevitably had to result in a lower birth rate through abortion, infanticide, and birth control (all of which he regarded as sinful) or in a higher death rate through war, famine, and pestilence. In one sense, Malthus was quite correct, since birth control became a widespread way to control population growth, to the point where the UN medium projection for world population shows a leveling-off and even possibly a decline (United Nations 2007). But, of course, attitudes have changed since his time; what Malthus regarded as a sin is now widely regarded as a sensible way to deal with a potential problem. Moreover, food production has increased far beyond the limits that Malthus thought possible.

Despite the seeming irrelevance of the original Malthusian model, neomalthusianism is in many ways the dominant discourse in the public debate on environmental issues. Indeed, Charles Kegley (1993: 140) took for granted that there was an ‘unabated deterioration of the global ecosystem.’ Neomalthusians argue that we are living on borrowed natural capital, that our ecological footprint is excessive (Wackernagel/Rees 1996), and that at some point the scarcities will become so acute that drastic solutions are inevitable. Paul Ehrlich (1968: 11) announced 40 years ago that ‘the battle to feed humanity is over,’ and later, Ted Gurr (1985: 51) feared that that overpopulation, exhaustion of nonrenewable energy sources, resource scarcity, and pollution would lead to a crisis of Western societies and more broadly of the whole global system, resulting in greater inequalities, more authoritarianism, and more widespread group conflict.

While Marxism tended toward technological optimism, today’s radicalism is to a large extent fused with neomalthusian thought. Thomas Homer-Dixon (1999) describes three forms of environmental scarcity: demand scarcity, supply scarcity, and structural scarcity. The first two relate closely to the original Malthusian model while the third refers to inequality.

While neomalthusian thought is very widespread, it is not unopposed. Cornucopians, or technological optimists, argue that a resource crisis can easily be averted by technological innovation, the substitution of resources, and market pricing. Attitudes change and environmental values start taking precedence over unrestricted economic growth once basic needs have been satisfied.Footnote 33 So, as the former Saudi Arabian Minister of Oil Sheik Yamani is reported to have said ‘… the stone age came to an end not for a lack of stones, and the oil age will end, but not for the lack of oil’ (Greider 2000). Of course, if the economic system is flexible enough to adjust gradually to the threat of resource scarcity, there is no need to fight over the scarcities. Therefore, ‘water wars’ and other major violence resulting from resource scarcity tend to be in the future. There is very little evidence for a general relationship between resource scarcity and civil war (Theisen 2008). Paradoxically, empirical studies show that the higher our ecological footprint, the more peace (Binningsbø et al. 2007). While a ‘water war’ rhetoric was very common 15 years ago, it has now largely been replaced by an emphasis on the need for cooperation in order to solve the very real problems of lack of clean fresh water in many areas of the world (UNDP 2006). As Riley (2001: 146) argues, Malthus may have been better at summing up the past than predicting the future.

Recently, the eco-war perspective has been revived in the debate about climate change. Climate change is indeed a very serious challenge. This is in part because of the accumulating evidence of probable physical effects of global warming, such as changes in precipitation, increasing sea levels, melting of glaciers, and increases in the number of violent weather events. Perhaps even more serious is the uncertainty associated with climate change. The numerous anomalies and deviations from the long-term trends illustrate the problem of making reliable short-term forecasts. They also make it very difficult to design policies for the prevention and mitigation of climate change.

It is evident that climate change will have consequences for human habitation, but using the physical models to derive social effects remains very difficult. It is not surprising, then, that projections for the social and economic effects of climate change are even more controversial than the physical effects. Moreover, while the IPCC summary of the physical effects are based on reviews of thousands of peer-reviewed studies in academic journals, the social effects rest on a much shakier scientific foundation. This is particularly true in assessing the possible security implications of climate change. NGOs, two successive Secretary-Generals of the UN, and numerous national politicians have surmised that climate change is a major security issue, and that we are now seeing the first of many climate wars in Darfur. But although climate change may have exacerbated the relations between herders and farmers in Darfur, area experts cannot disregard the policies of the Sudanese government, the ethnic and religious rivalries, the history of violence in the country, or the role of neighboring conflicts. As for the role of climate change in conflict more generally, there are very few peer-reviewed studies. Indeed, the IPCC (2007) reports are fairly cautious in commenting on the implications for armed conflict. But where they do, they rely on scattered and peripheral sources.

Had I been a neomalthusian addressing the issue of climate change and conflict 15 years ago, I might have been tempted to point out a disturbing covariation over time between global temperature increases and armed conflict, since both had been on the rise. In the most recent decade and a half, the variation is reversed; higher temperatures are associated with less conflict. However, we can deduce very little from two superimposed time trends, although much of the debate about the social effects of man-made climate change is phrased in such terms.

It is not surprising that the apocalyptic nature of the climate change debate should give rise to dystopias like Alan Weisman’s (2007) recent book The World without Us, where the author concludes that without man earth would be in good shape,Footnote 34 and even fringe phenomena like the Voluntary Human Extinction Movement (VHEM) and the Church of Euthanasia, with its four pillars of abortion, suicide, sodomy, and cannibalism. For social scientists, however, a more pressing issue is how earth cannot just survive with man, but even prosper.

4.4 The Commercial Challenge: The Capitalist Peace

The fourth and final challenge to the liberal peace comes from within the liberal school. One of its origins is the old observation that a democracy has never been established in a country that does not have a market economy. In other words, a market economy is a necessary, but by no means a sufficient condition for democracy.

Hegre (2000) has argued that the relationship between trade and conflict is contingent upon development. With increasing economic development, the cost of seizing and holding a territory increases, and the expected utility of conquest decreases. Developed states are therefore more likely to be trading states. Mousseau (2000, 2003a) found that both the democratic peace and the zone of democratic cooperation are substantially limited to economically developed democracies (see also Mousseau et al. 2003). Taking a further step backwards in the causal chain, Weede (1995, 2005) argues that economic freedom, of which free trade is an important component, promotes economic development and thus lays a foundation for democracy and for peace. Mousseau (2003b; see also Mousseau/Mousseau 2008) argues that it is the rise of contractual forms of exchange within a society that accounts for liberal values, democratic legitimacy, and peace among democratic nations. Gartzke (2007) maintains that the existence of market freedom accounts for the effects usually attributed to democracy and trade in analyses of the liberal peace. McDonald (2007) also finds that greater quantities of publicly held assets lead governments to pursue more aggressive foreign policies and increase the likelihood that they will participate in military conflict. Thus, he argues, capitalism promotes peace. Gartzke and Weede both use the term ‘the capitalist peace.’ There is some disagreement in the literature as to whether the democratic peace should be seen as a mechanism of the capitalist peace or as an alternative theory.

For someone who grew up with the idea that capitalism produces ‘merchants of death’ (Engelbrecht/Hanighen 1934) who profit from war, the capitalist peace is a difficult notion to swallow. But there would be little point to doing research if all the answers were given ahead of time.Footnote 35 I take refuge in the teachings of yet another former President of the ISA, Kenneth Boulding (1989), who distinguished between three forms of power: threat power, economic power, and integrative power. Threat power builds on force and the threat of destruction. Economic power rests on exchange and enlightened self-interest. Integrative power depends on legitimacy, respect, or even love. An actor does something not because he or she is forced to, or even because it is in his or her best interest, but because it is right. The family and many organizations depend mainly on integrative power; although, there may also be elements of force and exchange. Boulding argues that this is the most significant of all forms of power; and in the long run, neither threat power nor exchange power can be upheld without a minimum of legitimacy. Boulding described possible futures rather than predicted one that was most probable. But he certainly viewed a future where integrative power played the major role as the most hopeful scenario for mankind. In the decade and a half since Charles Kegley’s Presidential Address, the world has moved in large measure from a threat system to an exchange system. Perhaps in the next 15 years we can discern a clear movement in the direction of an international integrative system. Then, we can really speak of a neoidealist movement in international relations, and I take the liberty of invoking in support of this trend a popular slogan from my own youth: Make love not war! Meanwhile, even if love does not yet govern the world, most of us will probably be pleased that force has been replaced by commerce to such a large extent. Perhaps at this stage we have to make do with a less radical slogan: Make money not war!