In his seminal book Power Kills, Rummel (1997) summarizes decades of research on the democratic peace to make a single, pointed argument: that the worst kinds of violence—mass killings carried out by governments—are entirely explained by the tyrannical nature of the regime that commit such crimes.Footnote 1 His proposed solution to eliminating ‘democide’—as well as collective violence, war initiation, and other forms of political violence—is a well-known known one: to promote and reinforce democratic government. His view is that strengthening democracy, both in terms of procedural practices and qualitative, liberal behavior—could result in world peace, defined as eliminating violence between states and within them. Rummel concludes that democracy is ‘a method of nonviolence’—the subtitle of his book. In this chapter, I share some reflections on Rummel’s basic argument, offering some observations, some critiques, and some paths forward for contemporary research on peace, democracy, and nonviolent change. In particular, I highlight one key deficit of Rummel’s thesis: that he neglects the power of nonviolent civil resistance in bringing about democratization from below.

11.1 Rummel’s Motivation

What were Rummel’s stated intentions in writing this book? He overtly expressed his normative goals. Although his study offers a sophisticated and detailed review of the empirical patterns of democracy, democide, war, and collective violence, he clearly lays out the primary source of his interest—to reduce war in all its forms while also elevating what he saw as the most practical and realistic method of political rule. In the preface, he writes:

‘I hope to have something specific to recommend about ending war. But in the mid-1980s I was shocked to discover that several times more people were killed in democide (genocide and mass murder) by governments than died in warfare. And with that my aim broadened to help end or at least lessen this killing as well. This book presents the sum of all this research. And, I believe, I can finally offer what appears a most realistic and practical solution to war, democide, and other collective violence’ (1997: ix).

Later, he writes: ‘There is one solution to each and the solution in each case is the same. It is to foster democratic freedom and to democratize coercive power and force. That is, mass killing and mass murder carried out by government is a result of indiscriminate, irresponsible Power at the center’ (ibid: 3).

What strikes the reader is the somewhat rare and refreshing statement about the author’s own normative commitments—that of global emancipation as well as freedom from violence—motivating him to undertake the study. Such bold and self-revealing statements do not often appear at the outset of contemporary political science works, which are more typically characterized by the statement of an empirical puzzle, a correction to a theoretical framework, or the presentation of a new case. Here one sees an author who is not just interested in knowledge for its own sake, but who also wants to know how his academic discipline can bring its insights to bear on what he viewed as the most important questions of our time.

11.2 The Basic Argument

Rummel argues that democracy is indeed the most effective method by which societies can reduce violence. Although he concedes that institutional and cultural arguments have some merit, he also insists that his own field theory best explains the different norms and practices that vary across democratic and authoritarian regime types. In his view, democracy is comprised of social fields, constantly in flux and normalizing the bounds of civilized behavior; authoritarianism is static, increasingly rigid, and generating the cleavages over which people resort to violence to secure their survival (or prosperity).

Throughout the text, Rummel meticulously details the various arguments and empirical evidence for democracy’s generally pacifying effects. He is never naïve in his portrayal of these effects. He suggests that democracy reduces violence by degrees; that although imperfections persist, democracy remains the most peaceful form of government devised by human societies.

Recent empirical work has generally supported this claim. Indeed, notwithstanding a few definitional quibbles (e.g., Oren, 1995; Rosato, 2003; Ray, 2003), a major war between modern democratic states has still never occurred (Goldstein, 2011). Current studies confirm that civil wars are less likely in democracies (Hegre et al., 2001) and mass killings and democides are virtually unheard of (Davenport, 2007), although critiques abound as to the structural violence imposed by Western democracies on ‘periphery’ states as well as the ethical and normative implications of democratic peace theory (Galtung, 1990; Hobson, 2011). Terrorism, although somewhat common in democracies, has become something of a substitute for civil war in them (Chenoweth, 2010; Li, 2005), suggesting a decline in severity of political violence in freer states. Such patterns would square well with Rummel’s suggestion that democracy reduces most forms of violence, although it would be impossible to eliminate violence from society entirely. His main concern—and the main focus of his book—is on eliminating unrestrained state violence: a goal which he argues is only achievable through the robust construction of democratic societies.

As the book progresses, each chapter addresses common critiques of democratic peace theory on both theoretical and empirical grounds. For example, he was very concerned that the critics of democratic peace were overrating minor skirmishes and equating them to very destructive wars. In responding to the argument by Mansfield & Snyder (1995) that the democratization process can usher in particularly unstable and bloody periods within transitioning states, Rummel argued (p. 102) that Mansfield & Snyder supported this argument by looking at conflict frequency rather than conflict intensity—a crucial conceptual distinction. All conflicts are not equal, according to Rummel. And while transitioning states may be vulnerable to civil conflict, the lethality and global consequence of such internal wars pale in comparison to the far-reaching devastation wrought by major power wars in the twentieth century.

11.3 Nonviolence as the Absence of Unrestrained State Violence

Rummel conceptualized ‘nonviolence’ in its most elementary way—the lack of violence, and the use of alternatives to war that are not violent. For example, he expects that democracies in disagreement with one another will negotiate, since such forms of conflict resolution are deeply embedded in the culture, social expectations, and practices of democratic societies. He expects democratic leaders to eschew overreaction to domestic political opposition—and to avoid mass killings.

He does not, as far as I can tell, conceive of nonviolence as the active promotion of nonviolent contentious action, or as an active and coercive method of conflict in itself. Around the same time Rummel was completing his thesis, for example, a series of scholars such as Thomas Schelling, Gene Sharp, and Adam Roberts were writing about civilian-based defense—a sort of nonviolent deterrent to foreign occupation where civilians would train in civil disobedience so as to make any foreign occupation too costly and therefore unlikely (see, for instance, Roberts, 1969).

Nor does he conceive of nonviolence in the principled sense in which it is often proffered today (i.e. pacifism), where one would advocate the avoidance of violence simply on the grounds that it is immoral.

Instead, he views nonviolence simply as the absence of violence. As many note, a simple lack of overt violence is a far cry from a positive peace (Galtung, 1996). Rummel is not blind to this—he suggests that a lack of major violence is his goal, and that minor forms of violence will inevitably remain even within democratic states.

Of course, some critics aver that democracy reinforces and promotes structural violence in many different forms (e.g. Hobson, 2011). Rummel sees such critiques as emanating from a leftist bias from within academia, and he sees proponents of such views as influenced by socialists and Marxist critics of classical liberal theory more generally (1997: 100–115). By focusing on the failing of democracies and the structural violence perpetrated by them, he argues, such critics trivialize the scale of violence witnessed by war and democide. That said, he tries to address some of these critiques in his 2007 Blue Book of Freedom, in which he claims that political freedom provides economic and food security. However, he seems to miss a primary critique regarding structural violence—that economic, social, and political freedoms are unequally distributed even within ‘free’ societies based on race, gender, class, or other arbitrary social distinctions. Somewhat surprisingly, Rummel may have missed an opportunity in such dismissals—an opportunity to generate data to demonstrate the value of democracy on various other indicators of social life, such as economic equality, rule of law, quality of life, human development, trust in government, etc. This could have demonstrated that democracies perform well on a number of other indicators besides direct violence. Because his work is so empirically driven, it would make sense to allow empirics to shore up the case for democracy on these other counts as well.

Moreover, one can recognize the benefits of liberal theory without arguing that its applications have been flawless. For instance, the benefits of free societies have been distributed unevenly within them. Rummel himself brings up several examples of direct violence occurring in democracies, such as the beatings of Rodney King (and the subsequent race riots of 1992) in the United States. Such violence occurs quite easily and commonly in democracies. And this is precisely the concern of many critics of democracy and neoliberalism—that they conceal the worst kinds of abuses under the veil of a satisfied, apathetic, and perhaps privileged majority. That is, democracy and justice are not necessarily synonymous with one another, and justice may be a superior indicator of peace compared to democracy.

11.4 The Path to Democracy: Necessarily Bloody?

Rummel expressed frustration at the blunt measures at the disposal of states in bringing about democratic transition. He suggested that many forms of violent intervention, such as the sponsoring of violent rebellions, are anathema to the whole enterprise of both peace and democracy. In Power Kills, Rummel (1997: 9) clearly rejected military intervention as a method to bring about democracy in foreign countries. However, he later reversed course on this, suggesting that the invasion of Iraq in 2003 was necessary to stave off Saddam Hussein’s brutality and catalyze a wave of democratic transition transitions in the Middle East—and justified in its attempt to form an alliance of democracies to support the invasion (Rummel, 2005a; Tucille, 2014). His prediction was that although an American invasion and occupation would be temporarily painful, it would mostly be ruinous for the dictator and his inner entourage; in the end, the invasion would allow Iraqis the opportunity to liberate themselves from Hussein’s tyranny and obtain the power and opportunity to chart their own course forward.

Hence, through one of Rummel’s most controversial political stances, he found himself ideologically allied with neoconservative foreign policy elites in endorsing the United States’ invasion of Iraq in 2003 (and, indeed, at one point he endorsed censorship on the media on his blog; Rummel, 2005b). Like many neoconservative thinkers, Rummel proved to underestimate the devastating impacts of that war, in terms of both its immediate devastation and its continually bloody aftermath. Some commentators have even suggested that the war in Iraq initiated the period of turmoil in which the Middle East continues to find itself today (Tyler, 2015). It is impossible to know how Rummel would interpret this debacle now—and whether, in hindsight, he would have seen the Iraq War as worth the price in blood and treasure.

In the end, Rummel (1997: 9) suggests that plebiscites or referenda are the ideal ways for people to assert their power and choose their own government. Yet the implementation of this suggestion remains impractical. How are such plebiscites to come about in autocracies? This question is left unanswered. This leaves open the possibility of forced regime change, which has proved ineffective at best and disastrous at worst (Downes & Monten, 2013).

Perhaps he would have revised his views on Iraq if he had known about the realistic alternatives to violent conflict in bringing about democratic transitions. For instance, one wonders what Rummel would have made of the diffusion of mass nonviolent uprisings during the second half of the Twentieth Century—and into the current one (Karatnycky & Ackerman, 2005). Since 1970, the world has witnessed the explosion of mass upheavals characterized primarily by nonviolent forms of contention (Chenoweth & Stephan, 2014). Although states have used exceptional methods to try to counter these uprisings, we have nonetheless seen various waves of these mass movements in Eastern Europe in 1989, the former Soviet Bloc in the mid-2000s, the Arab Spring in 2011, and throughout the industrialized world since 2011. Data on these movements show that mass nonviolent contention has virtually replaced armed uprisings in terms of frequency (Chenoweth & Stephan, 2011). Despite what observers might infer from watching the news, mass violent rebellion is going out of style, and mass nonviolent contention is the new game in town—at least for the time being. As a consequence, nonviolence is now a prominent research agenda within political science (Chenoweth & Cunningham, 2013).

This recent upswing of nonviolent mass movements is not unrelated to Rummel’s core interests, and it may even connect to one of his remaining key practical puzzles—that of how countries actually achieve democracy through nonviolent means.

In an article on democratization, Ulfelder (2005) finds that one of the strongest associations is that between mass protests and subsequent democratization. Similarly, a report by Freedom House finds that 75% of recent democratic transitions were initiated in part (or in whole) by high levels of active civic pressure from below (Karatnycky & Ackerman, 2005). And campaigns of civil resistance—or nonviolence—are more likely to usher in democratic transitions than their violent counterparts (Chenoweth & Stephan, 2011). Indeed, in recent private correspondence, Jay Ulfelder has suggested that he now sees protests as a necessary (but not sufficient) condition in bringing about democratic transitions. According to two separate studies by Johnstad (2010) and Celestino & Gleditsch (2013), such nonviolence-initiated democratic transitions are exceedingly durable.

Rummel’s omission of nonviolent resistance from discussion is ironic on several counts. First, his field theory explicitly argues that democratic politics enjoy social fields, which involve ‘a high level of nonviolent conflict across the society, the stuff of democratic politics’ (Rummel, 1997: 147, emphasis in original). Indeed, nonviolent conflict—and nonviolent resistance in correcting democratic politics when necessary—is clearly linked to the key mechanisms through which Rummel argues that democracies are more pacific than authoritarian regimes. Rummel clearly knows and understands the literatures on nonviolent conflict, which he cites briefly in a footnote (1997: 102). Yet he totally sidesteps these literatures, seeing nonviolent resistance as an outcome of democracy rather than a cause of it. There is but one mention of Gandhi in the book, but only as a passing reference to charismatic leaders existing within social fields (1997: 156). Nor does Rummel appear to see constructive program (or parallel-institution-building) as a viable way through which societies can pursue models of democracy that suit their own interests. Rummel’s volume offers a people-powered explanation for the functioning of elite politics, yet he misses the most potent source of change witnessed in recent history: the people power movements that brought down many of the tyrannical governments he so deplored.

At the same time, Rummel underestimates the plural nature of authoritarian regimes—and the degree to which social fields exist in authoritarian regimes. The past several decades of scholarship on social capital, civil society, and authoritarian regimes, reveals that even in authoritarian regimes, people practice everyday forms of resistance, which can develop into collective outbursts of nonviolent civil disobedience at seemingly a moment’s notice (Scott, 1987; Kuran, 1991). Rummel’s omission is somewhat ironic, given that he recognizes that democratic governments are not monoliths either (Rummel, 1997: 17). In fact, in many of the authoritarian regimes he cites as having anti-field qualities (USSR, Iran, Chile), people power movements developed and challenged seemingly impenetrable regimes that had, indeed, ruled with excessive brutality (Chenoweth & Stephan, 2011). In other words, social fields may be a cause rather than a consequence of democracy.

11.5 Democracy as the Logical Conclusion?

One wonders what Rummel would think about democracy’s current challenges. After the fall of the Soviet Union, many observers averred that humanity had arrived at ‘the End of History’ (Fukuyama, 1992). Liberal democracy had won the ideological race, and all that remained was to transform the last remaining hold-outs of the authoritarian class age. This optimism about the pacifying effects of democracy became so widely accepted among political elites that it was informing nearly every national security strategy from the early 1990s onward (Miller, 2012). Given the obvious dividends of democracy, it was only a matter of time before all states voluntarily accepted its tenets—or, if not voluntarily, succumbed to these tenets by force. The inevitability of a global system of republics was taken for granted.

However, Freedom House (2015) indicates that 2014 was the 9th year in a row that aggregate democracy scores have declined. With very few exceptions, 2014 was a year of reversal for democracies, with serious backsliding in major global payers like Brazil, Turkey, and Russia, in mid-level states like Hungary, Venezuela, and Azerbaijan, and elsewhere. Given the fact that Rummel saw democracy as inherently superior—and that he viewed the logic of its pacifying effects as universally appealing—one wonders how he might explain these trends.

In fact, the future of democracy, justice, and nonviolence may be linked. Stephan & Burrows (2015) suggest that authoritarian backsliding has been occurring alongside the closure of space for civil society organizations. As authoritarian regimes wise up to the disruptive potential of people power, they try to crush such challenges in their infancy through various forms of smart repression (Chenoweth, 2015). Their attempts to do so are quite revealing about the types of power that truly threaten them.

Although I never met Rudolph Rummel, my guess is that he would have interpreted these trends as deeply troubling because of the constriction of social fields, and that he would predict war and violence as becoming more likely in the midst of such reversals. But because he underestimated the potential of people power movements, his skepticism may have been misplaced. There is a potential solution to the problem of democratization without bloodshed, and the problem of improving unfair and unjust practices within democracies as well: nonviolent resistance.

Indeed, the rise of mass nonviolent movements and their effects on systems of government worldwide may represent one final, unwritten chapter of Rummel’s book. He might have concluded that while democracy is a method of nonviolence, it may also be true that ‘Nonviolence is the method of achieving democracy.’