DEMOCRACY AND CONFLICT WITHIN STATES

In the political science literature, the term “democratic peace” almost always refers to the extraordinary infrequency of war and other forms of violent conflict between democratic states, as well as the body of theory and research explaining it.

Early work by political scientists derived from the fairly intuitive notion that people rebel when they feel deprived in some way. Deprivation and discontent breed anger, which may be combined with military means to provide the genesis for armed insurgency against the state. Feelings of deprivation are not necessarily based on objective conditions; they are relative to conditions that people believe they deserve but have been denied, due either to discrimination or incapacity on the part of the government (Davies 1962; Gurr 1970; Tilly 1978). Grievances against the government or other groups within society can be related to political conditions, economic conditions, or both.

More recent work, including research by economists, starts by observing that social discontent is seemingly ubiquitous and that demand-side explanations of civil war fail to account for the far less common occurrence of civil war. Among the many societies that are home to especially aggrieved groups, what distinguishes relatively stable societies from those experiencing civil war, according to this perspective, are the impediments to collective action, which limit the supply of rebels willing to undertake the hardships and risks of armed insurgency (Lichbach 1995).

Grievance and political openness

The role of democracy figures most prominently in political theories of internal conflict. Democratic governments are less likely to curtail individual liberties and are more likely to grant equal protection under the law; the absence of such rights and freedoms is frequently a source of resentment on the part of disadvantaged groups within nondemocratic societies. Where grievances do exist in democratic societies, the openness of the political system allows group discontent to be expressed nonviolently, including by way of confrontational but nevertheless lawful means like strikes and protests (Eckstein and Gurr 1975; Diamond 1999). Disadvantaged and oppressed groups in nondemocratic societies have relatively few alternatives short of open rebellion against the state. The proposition that democracies are at lower risk of civil war and other forms of mass internal violence has received some empirical support (e.g. Rummel 1995; Gurr 2000), but this finding is not as robust as the evidence for the democratic peace between states (Hegre and Sambanis 2006).

Supply side explanations found in many economic theories of civil war tend to downplay the significance of democracy because political rights and freedoms are assumed to bear upon the degree of grievance and discontent found within societies, fac­tors judged to be of secondary impor­tance. However, the extent that political systems vary in terms of the inclusive­ness of popular representation pertains also to the opportunity cost of rebellion. The decision to pursue extralegal means to redress grievances is a decision to forego the opportunities available within the political system.

One possible reason for divergent expectations regarding the role of democratic governance in reducing the probability of civil war is that theory and research on the question has not paid enough attention to the different types of civil wars and rebellions that societies experience. The factors explaining ethnic and nonethnic civil wars are not exactly the same. Ethnic civil wars pit ethnic communities against the state (and its supporters) and are fought over communal status within society. Nonethnic civil wars, like revolutionary wars, are fought over ideological or class-based disputes rather than identity issues. While economic theories of civil war see the opportunity costs of rebellion primarily in terms of economic opportunities, ethnicity and identity-based grievances are often unrelated to economic deprivation (see, for example, Rothchild 1997; Arnson and Zartman 2005). A distinguishing feature of ethnic civil war, according to Sambanis (2001), is resentment at the absence of civil rights and freedoms, discrimination in the adjudication of disputes between societal groups, and the lack of political representation or regional autonomy. These sorts of grievances emerge less often in democratic political systems, and Sambanis shows evidence that ethnic civil wars are indeed unlikely to occur in democratic societies. He also finds that if a state has democratic neighbors, it is less likely to become embroiled in an internal ethnic conflict.

The suppression of dissent is a defining feature of closed political systems. Therefore, for any given level of grievance (and other factors contributing to armed rebellion), we would expect authoritarian states to experience fewer civil wars than states with more open political systems. That is, the most democratic societies face few rebellions because the level of grievance is generally lower; group conflict is more often resolved nonviolently, even if sometimes contentiously. But the most authoritarian societies may also face few rebellions, despite a higher level of grievance, because group conflict tends to be suppressed by the state. This parabolic relationship between political openness and civil war is one that has been repeatedly reported in the empirical literature (e.g. Muller and Weede 1990; Ellingsen 2000; Hegre et al. 2001).

Where does this leave partially demo­cratic states? The literature does indeed suggest that these are the societies most likely to experience serious internal conflict. When political rights and freedoms are not fully respected, grievances emerge within disadvantaged groups. Although partially democratic systems, just like fully democratic systems, may permit the mobilization of groups whose interests are at odds with the policies pursued by the government, they are also quicker to repress protests, strikes, and other forms of civil disobedience, thereby inflaming discontent and resentment even as they close off opportunities for groups to redress their grievances. When a regime combines the permissive elements of democracy with the repressive elements of autocracy, the risks of violent domestic conflict are at their highest.

In many cases, partly free societies are also those undergoing political change. The argument linking civil violence to the transition from authoritarian to democratic rule is associated with the work of Samuel Huntington (1968, 1991), among others, and is the starting point for the contention, discussed above, that democratizing states are more likely to become involved in interstate conflicts.

Hegre et al. (2001) show that both of these factors—the institutional inconsistencies characteristic of partially democratic political systems and the volatility and opportunism associated with political change—are correlated with a higher probability of civil war. They estimate that partially democratic societies are four times as likely to descend into civil war as are complete democracies. Societies undergoing a regime change, whether in the democratic or autocratic direction, are at higher risk of civil war than stable political systems, and remain at higher risk for five years after the regime change. Although the findings suggest civil war is more common in partially democratic societies than in societies with autocratic governments, other studies have shown that autocratic governments are more likely than democracies to experience regime transitions (e.g. Gates et al. 2006). Thus, while democratization may bring a greater risk of domestic conflict, if democratic political institutions are fully consolidated, the new regime is at the lowest risk of civil war, due both to its institutional features and to its durability as a political system.

The impact of political openness and political change on the occurrence of civil war is called into question by some researchers because they posit that economic opportunity costs are a more important consideration than either group grievances or the opportunity costs of redressing those grievances through democratic institutions.

Settlement and peace-building

The conditions most conducive to preventing the outbreak of civil war are related to those most conducive to peace settlements and the reestablishment of political stability in the wake of civil war. The power of the central government must be consolidated, its legitimacy must be established or enhanced by allowing previously excluded groups access to the policy-making process, and sufficient economic resources must be mustered and allocated to support the peace-building process. The creation of each of these conditions may be assisted in various ways by external actors (Zartman 1995; Regan 2000). Intergovernmental organizations, for example, especially those composed mainly of democratic states, have been effec­tive in facilitating peace-building processes within both member and nonmember states (Pevehouse 2005; Pevehouse and Russett 2006). In the case of extremely destructive civil wars, and especially those in which group hatreds are acute and communal violence never far below the surface, real peace may not be possible in the near term. Political stability may require the establishment of a central government that rules with an iron fist rather than wide consent. However, draconian solutions of this sort fall outside the liberal paradigm guiding peace-building efforts since the end of the Cold War (Paris 1997).

Given the secondary role played by political practices and institutions in economic theories of civil war, relatively little attention has been devoted to questions of peace settlement and post-war reconstruction beyond highlighting the need to transform war economies (e.g. Pugh and Cooper 2004). An exception is Wantchekon's (2004) game theoretic analysis of “warlord democracy.” Wantchekon con­siders an interaction involving three players: two warring factions (warlords) and the citizenry. Each warring faction seeks political power and the economic benefits that come with it. While the civil war is in progress, each expropriates the wealth of a subset of the citizenry they control as well as mineral wealth of the territory they occupy. The interests of the citizenry depend on their affiliation with the contending warlords. Those who support one or the other enjoy both the warlord's protection and the full economic benefit of their labors and investments. Those who support neither receive no protection and the return on their economic activities is subject to expropriation. A continued state of belligerency, with its continued expro­priation of wealth, encourages unaffiliated citizens to invest less, which diminishes the warlords' take. Wantchekon suggests that both warlords may expect to do better by disarming and choosing democracy. In this case, each campaigns for the citizens' vote by proposing a tax rate, and the electoral competition brings promised tax rates down to the point where expected government revenue equals the warlord's take under conditions of continued belligerency. Because the legal tax rate is to be applied to the economic activities of the entire citizenry, it is lower than what is applied to the smaller popu­lation under each warring faction's control during belligerency. The winning faction's total take will be improved relative to the status quo, because citizens will now be investing more, and as long as both factions estimate that they have a reasonable chance of winning forthcoming elections—which Wantchekon shows need not be greater than one-half—democracy is the rational choice (Wantchekon and Neeman 2002; see also Przeworski 1999).

This notion of warlord democracy stands in contrast to the argument that political order in the aftermath of social upheaval may require authoritarian rule, or at least a substantial measure of illiberalism(e.g. Huntington 1968; Zakaria 2003). Still, although even those who once profited from civil war may benefit from a system of democratic governance, getting there is rarely a simple matter. Several studies have pointed to the difficulty of post-civil war political reform and the fragility of democratic institutions when the wounds of communal conflict are fresh and the process of national reconciliation is in its infancy. Power-sharing arrangements can be essential elements in bringing warring factions to a settlement by assuring the representation of group interests in policymaking and state-building (Walter 2002; Hartzell and Hoddie 2003). But members of a governing cartel of group elites may resist the consolidation of more participatory political reforms as hard-won gains become jeopardized by the uncertainties of the democratic process (Rothchild and Roeder 2005). Majoritarian solutions have a mediocre track record in pluralistic societies plagued by suspicion and hostility; the absence of guarantees leave minorities understandably fearful of demagoguery and democratic tyranny.

In some post-conflict environments, there is such a thing as too much democracy too fast. Stable democracy provides mechanisms for the peaceful resolution of social conflict, but democratic governance (like market capitalism) is inherently competitive. Electoral competition, perhaps more than any other single institution, is emblematic of the democratic process, which is why the holding of elections after peace settlements have been reached is so often taken as an indication that a previously dysfunctional polity has made a successful political transition. In a pluralistic setting, however, electoral competition may encourage candidates for office to distinguish the interests of their constituents from those of other groups, and even to exaggerate them, reinforcing the perceived group differences that fed the violent conflict from which society just emerged. This does not mean that elections should necessarily give way to other, less participatory forms of political representation, like power sharing, but it does mean that electoral rules need to be designed so that candidates are compelled to seek support from outside their own factions. Extremist appeals should not pay on election day. It may also mean that elections cannot be rushed, that they should be delayed until such time as violent passions have subsided, warring factions have been disarmed, and the social infrastructure has resumed its basic functions.

The danger, of course, is that the difficulties associated with peace-building and state-building will become a justifi­cation for delaying indefinitely democratic consolidation. In the contemporary scholarly literature, there is a near-consensus that democratic political reform is a necessary component of peace-building after civil war, at least once a modicum of post-conflict stabilization has been achieved. Those critical of prevailing practices of “liberal peace­building” usually do not question the ultimate aim, but rather the rate at which the political process is opened up to societal forces or the design of representative institutions in the near term (e.g. Paris 1997, 2004; Fukuyama 2005; Barnett 2006).