Democratic Reciprocity*

A. Positive and Negative Reciprocity

Reciprocity is a self-limiting norm. It is not an escalating disposition, but an equilibrating one. It aims at balancing social relations, not at stretching charity or heightening hostility. Janus-faced, it encourages cooperation and permits conflict, while limiting both. It demands bounded (“appropriate”) responses to kindness, as well as to nastiness. Both the retribution of benefits received and the retaliation of harms suffered must be qualitatively adequate (“fitting”) and quantitatively commensurate (“proportional”).5

B. Sustaining and Shifting Equilibria

In democratic politics as well as in social life, norms of reciprocity are “equilibrium preserving dispositions.”6 They preserve social equilibria of productive exchange,7 as well as democratic equilibria of fair cooperation.8 Yet, given their two-sided nature, they serve to sustain cooperative as well as conflictive equilibria. As long as others cooperate, reciprocity demands cooperative responses. When others fail to cooperate, it grants permission to do the same and refuse cooperation. Hobbes’s state of nature, in which life is “solitary, poor, nasty, brutish, and short,” is as much a product of social reciprocity as ancient friendship or modern romantic love.

However, while helping to preserve social equilibria, reciprocity, too, carries the potential of altering equilibria as soon as “deviant actors” contravene established patterns of interaction. When people start cooperating in hostile environments, norms of reciprocity may engender cooperative responses in others. When they stop cooperating in cooperative environments, these very same norms may create spirals of retaliation that obliterate all cooperative dispositions. In responding to anomalous acts of cooperation, the power of reciprocity is restorative; it may lead actors back to cooperative equilibria. In responding to unilateral acts of conflict, its power is corrosive or even disruptive; it may destroy cooperative equilibria.

C. Reciprocal Democratic Cooperation

Common concepts like “mutual security,”10 “mutual guarantees,”11 “mutual respect,”12 “mutual toleration,”13 and “elite convergence”14 express the same idea of democracy as a cooperative system of reciprocal constraint. Moreover, all established definitions of “democratic consolidation” or “democratic equilibria” require reciprocal compliance by all relevant actors. Some demand more (e.g. widespread legitimacy), but none less. None permits political actors to cheat their adversaries or calls for self-sacrifice by their victims. All assume that actors participate in competitive elections, accept electoral outcomes, and refrain from coups and rebellions because their contenders do so as well.15

Strong assumptions of universal, reciprocal norm compliance have also been central to normative democratic theory in the tradition of John Rawls. The “fundamental organizing idea” of his theory of justice as fairness is the idea of “society as a fair system of cooperation over time,”16 which is grounded in the principle of reciprocity: citizens accept its rules and principles “provided that everyone else likewise accepts them.”17 Accordingly, his “ideal” theory of institutional justice assumes “strict compliance”18 by all citizens with their “natural duty to uphold justice.” “Everyone is presumed to act justly and to do his part in upholding just institutions.”19 In his perspective, “reasonable” citizens can be expected to comply “willingly”20 with their “obligation to play by the rules,”21 “given the assurance that others will likewise do so.”22

D. Reciprocal Democratic Defection

Alongside the reciprocal foundations of democratic stability, empirical scholars of democracy have stressed the reciprocal sources of democratic breakdown. In typical processes of democratic crisis and demise, norms of “positive reciprocity” crumble, while practices of “negative reciprocity” escalate. Someone starts ignoring democratic norms, others follow suit, and before long they are all caught up together in “an unending string of mutual defections.”23 Once the bonds of positive reciprocity are broken, democracies risk disintegrating in “an escalating tit-for-tat”24 of retaliatory norm violations. For instance, the processes of democratic breakdown in interwar Europe have been described as processes of reciprocal escalation of violence. Both extremes of the political spectrum grew armed wings, attacked each other, carried out political assassinations, entrenched themselves in perceived security dilemmas.25 Similar dynamics of reciprocal political violence and threat perceptions drove the Latin American wave of democratic breakdown in the wake of the Cuban revolution.26

So, clearly, norms of reciprocity carry the twin potential of either sustaining or destroying democratic equilibria. What are the normative implications of their two-edged empirical nature? In normative terms, the demands of “positive reciprocity” that direct citizens to “endorse and act on” democratic principles, “provided others can be relied on to do the same,”27 are unproblematic. In ideal situations of mutual cooperation, they ensure fair relations among actors as well as fair institutions. In “non-ideal” situations of only “partial compliance” with basic democratic rules, by contrast, normative tensions arise between the demands of “negative reciprocity” and democratic stability. Given the corrosive potential of negative reciprocity, democrats cannot embrace it in the same unambiguous manner as positive reciprocity.

A. The Risks of Retaliation

When transgressive actors violate the cooperative ground rules of democratic competition, their rule-compliant adversaries “can justifiably feel anger and resentment.”35 The norms of negative reciprocity invite them to retaliate and respond in kind. Negative reciprocity protects the fairness of competitive relations. It permits political actors to punish their transgressive adversaries and guard themselves against abuse and exploitation. It allows them to make their cooperative dispositions conditional on the good behavior of others and to protect themselves against unilateral aggression. In the face of foul play, they need not suffer in silence and “cede the political realm to the least scrupulous actors.”36 While norms of “positive” reciprocity oblige them to reciprocate the fidelity to democratic norms others show, norms of “negative” reciprocity permit them to retaliate against normative transgressions others commit.

The upside of direct retaliation lies in its capacity to restore the scales of justice (“an eye for an eye”) and to protect the competitive standing of injured parties (as both sides end up one-eyed). Its downside lies in its potential to destroy the cooperative foundations of democratic competition.

When democratic contenders play rough (within rules), rather than dirty (against the rules), direct retaliation often appears to be an appropriate strategy. Playing soft when others go hard, would be self-damaging. Competition is about winning. As long as competitive struggles remain within the bounds of basic norms, there is no need to go easy on competitors or pity the losers. Renouncing aggressive strategies which others deploy would amount to self-sacrificial “unilateral disarmament.”37 Retaliation may also be a legitimate response to normative breaches below the level of fundamental democratic norms. For instance, when an opposition party sheds informal norms of cross-party cooperation in favor of an uncompromising policy of legislative obstruction, their adversaries may be forgiven if they “mimic the same kind of ruthless obstructionism”38 once their roles switch.

In the face of violations of basic democratic norms, however, simple retaliation tends to threaten the institutional foundations of democracy. It does so in direct fashion: just as returning evil with evil does not create goodness but duplicates evil, answering democratic norm violations with further violations doubles the damage to democracy rather than repairing it. Retaliation also tends to inflict indirect damage on democracy by feeding spirals of punitive normative breaches. If one party starts, say, buying votes, libeling critics, or controlling courts, and others respond in kind, there is no natural stopping point to the escalation of mutual aggressions. In democracy, as well as in other social spheres, simple retaliation risks dissolving cooperative equilibria in “an unending echo of alternating defections.”39

In standard game-theoretic accounts of democratic equilibria, mutual threats of maximum retaliation (yet not retaliatory acts themselves) serve as stabilizing devices. If someone were to break the rules, their adversaries will go to war. Democratic peace rests upon the threat of armed conflict. It is an equilibrium of mutual dissuasion in which partial compliance does not exist. In these equilibrium models of mutually assured destruction, reciprocal threats of retaliation ensure strict and universal democratic compliance for as long as they work. Once they fail, however, actual retaliation, be it in the form of rebellion or military coups, ensures swift democratic breakdown.40

Now, if negative reciprocity in its simple form of direct retaliation serves valued goals (punitive justice and competitive self-protection) while threatening others (the protection of democracy), we need to weigh these “conflicting imperatives”41 and set priorities. In actual democratic politics, victims of foul play may wish to give free rein to their desire for retaliatory self-defense, even at the risk of subverting the democratic system. Against their single-minded quest for just retaliation, the idea of “democracy-preserving reciprocity” limits the demands of negative reciprocity in the name democratic protection.

Democracy-preserving reciprocity does not and cannot dissolve the tensions between punitive justice and competitive self-defense, on the one hand, and cooperative regime defense, on the other. Rather, grounded in a consequentialist Weberian “ethics of responsibility,” it strives to balance these goals in a way that gives priority to the preservation of democracy. It rests on the “principle of democratic responsibility,” which obliges actors to maintain “a sense of the damage likely to result from [their] defensive action.”42 Blending normative sensibility with strategic intelligence, it obliges them to heed the norms of reciprocity in ways that safeguard the entire system of reciprocity. Yet, if the easy route of direct retaliatory acts is closed off to them, how else can responsible democrats confront democratic norm violations?

B. Self-Defeating Passiveness

Given the destructive potential of symmetric retaliation against fundamental normative breaches, democratic actors sometimes consider its opposite: passive endurance. Refusing to hit back, to betray their commitments, to break the rules, they endure antidemocratic attacks in silence. They maintain the high ground of democratic morality. Instead of repelling aggression with aggression, they turn the other cheek. Rather than “going dirty” on rogue competitors, they adopt the heroic posture of “noble partisanship”43 and keep playing by rules their adversaries have chosen to abandon.

Again, silent suffering may be a plausible response to transgressions below the level of fundamental democratic norms. For instance, when a head of government (no need to name names) abandons all rules of decency in his treatment of political adversaries, insulting and ridiculing them like a resentful teenager, the victims of his rhetorical abuse may well choose to remain silent rather than feeding the news cycle with their response. In the face of basic democratic norm violations, however, the quiet heroism of unilateral “forbearance”44 tends to be “counterproductive.”45 While its spirit of self-sacrifice is laudable, it tends to achieve little more than self-sacrifice. By renouncing punitive justice, passive endurance avoids the mutual destruction of democracy in spirals of retaliation, yet opens the door to its unilateral destruction by freewheeling norm breakers.

C. The Virtues of Accountability

Rather than permissive endurance, democratic norm violations demand punishment. To re-establish cooperation, democrats must be prepared to enforce cooperative norms through disciplinary action. In Robert Axelrod’s terminology, they must show themselves to be “provocable.”46 If they turn a blind eye to democratic transgressions, they invite transgressors to repeat and deepen them. However, if they wish to avoid both the interactive and the unilateral demolition of democracy, their strategies of enforcement must steer between the Scylla of mechanical retaliation and the Charybdis of passive endurance. They can do so by mobilizing the variegated toolset of democratic accountability.

The first and basic principle of democracy-preserving reciprocity in the face of attacks against democracy is simple: fight them with democratic means. Seek democratic solutions to democratic problems. Call transgressors to account. Demand answers and seek punishments within democratic institutions. Counterattack by mobilizing any of the political resources democracy offers: speak out, articulate criticism and demand correction, get together with like-minded people, listen and talk to the other side, set up a petition, take to the streets, stage a sit-in, activate the media, tweet with prudence, join a party, campaign for your candidate, run in elections, initiate recall or impeachment procedures, stay informed, watch the legislative process, write your deputy, take your case to the courts, and so forth. As long as at least some democratic safeguards are intact, these tools of accountability provide the primary channels of negative democratic reciprocity.

Democratic accountability serves many purposes. It helps to expose, punish, redress, and discourage normative breaches. And in doing so, it reinforces rather than debilitates the cooperative foundations of democratic conflict. Its manifold pathways of “negative reciprocity” offer mixtures of punitive justice and competitive self-defense that respect and protect democratic ground rules. The quest for accountability may be regarded as a minimal democratic response to breaches of democratic duties. Democrats should not ask for less. But can they ask for more?

Do the principles of democracy-preserving reciprocity impose further restraints, or ease certain restraints, on political actors? I think they do both. In conditions of democratic normality, the imperative of preventing the emergence of retaliatory spirals imposes a set of positive duties on democratic actors. In contexts of democratic subversion, by contrast, the asymmetric balancing of justice, self-protection, and democratic defense opens the door to exceptional breaches of basic rules. It gives democratic actors license to resort to measured normative transgressions in the name of democratic resistance.

A. Resistance to Authoritarian Opposition

How should democratic states treat undemocratic actors? The theory and practice of self-defensive or “militant” democracy foresee a broad range of legal restrictions democratic states may place on the political participation of antidemocratic actors. They range from selective bans on party formation and electoral participation to extensive prohibitions on political discourse and behavior.61 Given the paradoxical nature of these restrictions, which strive to defend democracy against its enemies by measures which are inimical to its spirit, the legitimacy of democratic militancy has been subject to intense debates about its decision criteria. Is it legitimate for legislative restrictions to target antidemocratic ideologies or do we require actual harm to be done before we mobilize the force of repressive laws? Is it legitimate to suppress minoritarian actors who may have the intention, but not the capacity, to damage the democratic process? And is it legitimate to act in a preventive manner and ban or silence antidemocrats who do not pose “a clear and present danger” to democracy, but may come to do so in some indeterminate future?62 Beneath these inconclusive controversies lies a basic consensus. When political actors refuse to play by the rules of democracy, democrats may be entitled to place them under close supervision, limit their liberties and rights of participation, or even expel them from the democratic game.

B. Resistance to Unfairness

How should democrats treat unfair opponents? With Adam Przeworski we can understand democracy as “a system of processing conflicts” that “generates winners and losers.”63 Democratic allocations of gains and losses are legitimate insofar as they result from fair rules and procedures. However, complaints about unfairness are endemic in real-world democracies. As any casual observer of democratic politics knows, political contenders regularly violate, or are seen to violate, basic norms of fairness. They commonly do so within the law and well beyond the purview of militant democracy. They lie, deceive voters, spread false rumors, run dirty campaigns, play friends with media moguls, design electoral rules to their advantage, exploit legal loopholes to fill their party coffers, and so forth. And they do all this without breaking the law, propagating authoritarian ideologies, inciting violence, or posing an existential threat to democracy.

What should democrats do in response to such “hardball tactics”64 by their opponents? Reciprocating unfair practices runs the risk of generalizing them. When aggrieved democrats try to get even by hitting back, by paying back injuries with injuries, they can count on the applause of their supporters and the comprehension of their ideological neighbors. Yet, by answering unfair moves with unfair countermoves, they likely contribute to legitimizing, deepening, and entrenching political unfairness. To avoid the downward road of negative reciprocity, they may adopt the opposite policy of principled, unconditional fairness. As Michelle Obama famously formulated at the 2016 Democratic National Convention, “when they go low, we go high.”65 Bagg and Tranvik plead for a compromise: democrats must not “cede the political realm to the least scrupulous actors.”66 In the face of rogue competitors, they need to protect “those things that are truly necessary to democracy”67 but have license to reciprocate their adversaries’ “deceptive practices”68 with “mild forms of dishonesty.”69

An alternative route is democratic ostracism. Democratic norms demand that political adversaries recognize each other as legitimate participants in the political process. Democrats, however, may withdraw such recognition from contenders who systematically flout democratic norms. They may signal their principled disapproval by disengaging from them through passive silence or active isolation. In a policy of silence, they deny them attention and stop both listening and responding to them. In a policy of isolation, they draw a “political cordon sanitaire” around them and keep them out of public debates, governing coalitions, and other positions of power.70

C. Resistance to Electoral Manipulation

All accepted definitions of democracy—indeed, any plausible definition of modern democracy—include elections as its core institution.71 By implication, all accepted and acceptable catalogs of democratic norms include the demand that actors recognize elections as the only legitimate pathway to the peaks of state power. Losers are obliged to concede defeat, assume the role of peaceful opposition, and place their bets on subsequent elections. “If democracy is to exist, at least one rule must be observed, namely [that] political parties competing in elections obey their results.”72

However, both the legitimacy of electoral victory and the obligations of electoral defeat are not absolute, but conditional. They depend on the democratic nature of the elections. While elections are essential ingredients of liberal democracy, not any theatrical performance does the trick. Unless elections fulfill the minimum standards of liberal democracy—that is, by conventional phrasing, unless they are competitive, free, and fair—they neither legitimize the victor, nor put the losers under any moral obligation to consent to their defeat. Democratic “opposition parties are supposed to accept the uncertain outcomes of democratic elections, yet not the prearranged outcomes of authoritarian elections. In electoral autocracies, losing parties have no normative reasons to accept the verdict of the ballot box.”73

Again, mechanical reciprocity does not provide plausible remedies for authoritarian elections. The appropriate democratic response to electoral manipulation is not counter-manipulation. Rather than restoring democracy, mutual electoral fraud, “competitive vote buying,”74 or “competitive clientelism”75 further undermine it. Aside from mobilizing voters, the defensive repertoire of democratic opposition actors is largely limited to Albert Hirschman’s strategic pair of “exit” and “voice,”76 that is, the withdrawal of active cooperation (electoral and legislative boycotts) and the withdrawal of discursive support (public protest, judicial appeal). As Juan Linz writes, “when they feel that there is no opportunity for fair and open competition in the elections,” “even system-supporting parties” may resort to boycotting elections and parliaments and launching disruptive strategies of “strikes or mass protests.”77 Whether they participate or not in authoritarian elections, and whether they accept or reject their results, does not reflect their democratic commitments, but their strategic calculations.78

D. Resistance to Judicial Partisanship

How should democrats deal with partisan courts? The democratic demand that citizens accept the legal settlement of disputes by impartial courts presupposes that courts actually do what they are supposed to do: settle disputes in an impartial manner within the bounds of law. Judicial impartiality, however, is a precious, fragile good that can be destroyed in many ways. Potential sources or agents of partiality are manifold. Governments, governing parties, or “cartels” of political parties79 may subject courts to political control. External actors like private companies, religious authorities, or criminal organizations may capture them. Illicit currencies like money (corruption) and violence (intimidation) may override their legal reasoning, and internal ideological biases (such as racism, class prejudice, or machismo) may do the same.80

How should democrats deal with biased legal institutions? What are licit remedies against judicial actors who systematically violate their duties of impartiality? A first line of defense is procedures of oversight and discipline within judicial systems. But what if these internal “protections against judicial aberrance”81 fail?

At the very least, courts of justice should be prepared to subject their judgments to the court of public opinion. They should be willing to listen to critical voices, not just from those they condemn, but also from judicial experts, mass media, political parties, judicial activists, and the general public. “Criticism is not only inevitable but healthy.”82 Still, given the relative fragility of “the least dangerous branch,” both political and academic observers often perceive criticisms of judicial rulings by powerful actors as illegitimate attacks on judicial independence.83 Similarly, they tend to classify “populist” opposition actors who denounce the partisan nature of judicial agents as principled foes of constitutional democracy and its architecture of checks and balances.84

In the face of persistent judicial biases that prove resistant to criticism, democrats may go a step further and demand changes in the composition of judicial bodies. In the aftermath of dictatorship, “transitional justice” policies often strive to ensure institutional impartiality either by removing public officials of the ancient régime from key positions (“purges”) or by blocking the access of former authoritarian elites to state power (“vetting”).85 Broad, post-transitional lustration policies are bound to be controversial. The same is true of their targeted counterparts during ordinary democratic times. Alleged attempts to correct alleged political biases of judicial actors are almost invariably seen as attempts to introduce such biases into the judiciary. Except in the US, where parties and presidents have come to openly celebrate their success in colonizing the judicial system,86 democratic governments that try to alter the composition of courts usually do so in the name of justice, not under the banner of personal or partisan manipulation. If courts have been “packed with party loyalists,” they contend, restoring judicial impartiality obliges democrats to “unpack” them.87 Yet, their allegations of partisan bias, as well as their replacements of partisan judges, are themselves likely to be seen as partisan.

Whenever governments remove sitting judges or force them to resign or retire, they are suspected of political manipulation.88 For example, after its return to power in 2015, the Polish Law and Justice Party (PiS), which had for long “demonized judges as unreconstructed Communists,”89 carried out far-reaching judicial reforms with the official aim of “overhauling a corrupt system that obstruct[ed] popular will.”90 Unsurprisingly, applause came only from its own camp. Both its opponents and independent observers saw a partisan campaign “to take over the justice system,”91 a multilevel effort to “undermine the separation of powers [and] the independence of the courts.”92

On the other side of the Atlantic, in the US under Donald Trump, combative Democrats claim that Republicans have been entrenching the power of a social and ideological minority at the peak of the judicial system by capturing the US Supreme Court through “outrageous partisan warfare”93 and “democratic sabotage.”94 In the name of democratic salvation, they plead for a “judicial counterrevolution,”95 which involves a transformative Democratic “assault on the Republican Court majority”96 through court expansion, the end of lifetime tenure, and new appointment procedures.97 Their Republican fellow-citizens, however, are unlikely to embrace the underlying distinction between “constitutional hardball in the service of democratic erosion and constitutional hardball in the service of democratic preservation or renewal.”98 The democratic legitimacy of judicial purges hinges on the quality and persuasiveness of arguments about the partiality of sitting and the impartiality of incoming judges. In contexts of political polarization, such assessments of judicial partisanship tend to be irreconcilably partisan themselves.

E. Resistance to State Repression

How should democratic citizens respond to repressive governments? In principle, citizens have an obligation to obey the law. International law holds this obligation to be absolute,99 political philosophers tend to make it conditional on the justice of the political system.100 Generally speaking, a democracy is entitled to demand obedience from its citizens for as long as it remains a democracy. Under dictatorship, citizens may invoke their ancient “right of rebellion.” Yet, what are they entitled to do in ambiguous situations of selective repressiveness? How should they resist an illiberal government which, while perhaps remaining minimally democratic, turns to (either illegal or legally camouflaged) political repression without possibility of legal redress? Is there a legitimate third way between strict obedience to a democratic political order and rebellion against tyranny?

The classic response reads “civil resistance” or “civil disobedience.” In John Rawls’s seminal formulation, civil disobedience is “a public, nonviolent, conscientious, yet political act contrary to law,”101 that aims at rectifying a “persistent and deliberate violation”102 of political rights and liberties by appealing “to the sense of justice of the majority (or those having effective political power).”103 For citizens torn between their “duty to comply with laws enacted by a legislative majority (or with executive acts supported by such a majority),”104 on the one hand, and “the right to defend [their] liberties and the duty to oppose injustice,”105 on the other, civil disobedience offers a counter-strategy of “last resort.”106 When democracy is in danger, Juan Linz wrote in concordant manner, the “use of mass pressure by trade unions, tax payers, or citizens in the form of strikes or mass protests disrupting the operation of government” may be “the last recourse for defense of the system.”107

Of course, both the empirical realities and the moral grammar of civil resistance often diverge substantially from Rawls’s idealized conception. Often, campaigns of civil disobedience struggle against authoritarian rule, rather than striving to overturn democratic decisions, and are more contentious and transgressive than idealized accounts of civic self-restraint would allow for.108 Still, “the complexity of civil disobedience”109 does not invalidate the general idea that specific transgressions of the law may be justified to defend or restore the democratic legal order. When severe breaches of the law by public authorities undermine individual rights, including the right to effective judicial self-defense, citizens’ obligations to the law do not dissolve entirely, but loosen. Withholding their cooperation from state authorities without engaging in violent confrontation110 seems a reasonable compromise between blind obedience and blind rebellion. Still, the outer bounds of permissible transgressions in the name of democratic defense are less clear-cut than ideal-typical notions of civil disobedience suggest.

F. Resistance to Societal Violence

The contingent, reciprocal nature of democratic commitments is easiest to discern in situations where these commitments dissolve because their underlying conditions disappear. At its core, the idea of democratic reciprocity is simple: democrats play by the rules of the democratic game as long as others do—that is, as long as the democratic game is a democratic game. Democratic norms of fair play presuppose a fair game to exist in the first place. They are valid within the game, not outside. They are not norms of self-sacrifice or exploitation, but norms of cooperation which lose their binding force to the extent that their constitutive cooperative conditions cease to exist. The dependence of democratic norms on democratic conditions holds even for the primary categorical imperative of liberal democracy: the renunciation of violence.

The democratic rule of law imposes a universal ban on violent conflict resolution. It obliges citizens to settle their private, as well as their political, conflicts without recourse to violence. And it obliges the state to make this ban effective and provide for remedies (penal justice) in case of exceptional breaches. Again and again, scholars of democracy have stressed the intrinsic link between democracy and the general renunciation of violence. John Keane’s succinct formulation is representative: violence is “the greatest enemy of democracy as we know it. Violence is anathema to its spirit and substance.”111 The democratic prohibition of violence seems absolute. We tend to perceive those who reject it, and even those who doubt it, as enemies of democracy. “When the chips are down, the divide [between the friends and foes of democracy] is violence.”112

Despite the strong normative consensus about the democratic ban on violence, real-world democracies vary dramatically in the extent to which they control either political or private violence by social actors. Some democracies fail to impose the democratic accord of peaceful conflict settlement among political actors. They are challenged by organized violence from religious ideologues, as in Pakistan and Tunisia, left-wing insurgents, as in Colombia and India, or secessionist movements, as in Mali and Sri Lanka (until the military annihilation of the Tamil Tigers in 2009). Other democracies fail to impose the civil accord of peaceful conflict resolution among private actors. They are challenged by criminal violence from armed groups, such as the Red Command (Comando Vermelho) and the First Capital Command (PCC) in Brazil, and the Mara Salvatrucha (M-13) and Barrio 18 in El Salvador.113

What follows from the “horizontal subversion” of democracy114 through societal violence? Does the irruption of violence into their political or civil lives release citizens from their democratic duty of non-violence towards their fellow-citizens? If so, at which point and to what extent? In violence-ridden democracies, victimized actors often hold that direct counter-violence is justified. The retaliatory logic of “negative reciprocity” leads them to simply hit back. Members of rival communities engage in cycles of revenge and retaliation. Lynch mobs stage public assassinations of suspected violent criminals. Political parties resort to revenge killings, street violence, the organization of armed wings, or the delegation of violence to proxies. We may be able to comprehend the “just anger” of aggrieved parties that drives counter-violence, their “enough is enough” after some threshold of patience and tolerance is crossed. However, while sympathetic to calls for justice, the moderating principle of “democracy-preserving reciprocity” cannot give license to violent revenge. It must seek middle ground between passive suffering and retaliatory violence.

Again, the first line of democratic resistance reads accountability. The democratic default position is to keep relying on the institutions of law as far as possible. As long as existing institutions of security and justice offer reasonable prospects of protection and redress, we must constrain our impulses to take either security or justice into our own hands, and entrust our fate, as well as that of our violent aggressors, to the police and the courts. Still, at some point, the universally recognized right of self-defense comes into force, and allows us to meet threats of violence with threats of violence. At what point, though? What are the “reasonable” minimum prospects of public security and justice we demand for submitting ourselves to the rule of law rather than preparing for war? What are the thresholds of insecurity with impunity that propel us to take defense into our own hands? The most common and least controversial answer relies on human common sense: the intuitive perception of “clear and present” dangers to our physical integrity.

However, in the presence of immediate, indubitable threats of violence, democratic citizens’ right to individual self-defense is “relatively unproblematic.”115 But what about collective self-defense? The standard response of “defensive democracy” to “clear and present” threats of political violence is to ban prospective perpetrators and subject them to criminal prosecution.116 In quasi-anarchical contexts, however, the state lacks the capacity to do so. The case for collective self-defense arises from the systemic vulnerability citizens suffer in environments of pervasive violence by criminal co-citizens.

A similar general logic of self-defense against “horizontal” aggressions holds for non-political violence as well. Democracy is a political regime that regulates access to state power and its exercise.118 It presupposes the existence of an effective state.119 In its absence, when social order breaks down and citizens descend into a Hobbesian “state of nature,” private attempts at restoring order by local warlords or self-defense forces may be normatively justified. In such extreme contexts, “the normative priority of restoring public order amid generalized insecurity”120 may trump citizen duties to the nominally existing democratic state.121

The notion that quasi-anarchic conditions release democratic citizens from their duties of peaceful conflict resolution is essentially defensive. Striving to reconcile the goals of physical survival and democratic restoration, it permits the “pre-emptive” and “defensive” organization of private violence, but does not license unilateral aggression or violent retaliation. When we come to live under the shadow of “uncivil violence”122 by our fellow-citizens, we need not go like sheep to the slaughterhouse. We have permission to revise our commitment to legality and non-violence, though we need to do so without destroying the bases of democratic coexistence in the future. The idea of collective self-defense offers a precarious normative compromise. In the face of a dissolving monopoly of public violence, it allows the private organization of violence, yet limits its purpose. To avert a descent into spirals of reciprocal barbarism, private counter-violence is meant to serve no other end than protecting citizens against pervasive threats to their physical integrity.

The private organization of violence in the name of collective self-defense is an extreme measure in extreme situations. As a matter of course, its potential for abuse is immense. Almost invariably, the perpetrators of political violence in democracy speak the language of victimhood and self-defense. Lynch mobs pronounce themselves angels of justice, paramilitary forces commit atrocities in the name of public tranquility, ethnic rioters repair historical grievances by expelling friendly neighbors, and all of them present their aggressions as defensive acts against hostile forces that threaten their peaceful communal lives.123 Organizations of terror, death, and piety like al Qaeda and ISIS have described themselves as victims of imperial aggression and cultural hostility (“Islamophobia”) who seek retributive historical justice through indiscriminate murder (“persecution is worse than killing”).124 However, even absurd justifications of pre-emptive, defensive, or retributive violence, based on surreal notions of grievance and retaliation, or of threat and prevention, underline the irritating point: our unshakeable democratic commitments to non-violence are shakable. The democratic ban on violence is not absolute, but conditional; it flourishes in peace but wilts, self-defensively, under the shadow of violence.

In sum, when democracy dies, it takes some democratic commitments to its grave, but not all of them. Democrats need to fight their enemies in a self-restrained manner that permits their readmission as adversaries in a future of democratic restoration. Democratic reciprocity in the face of societal violence involves yet another balancing act: admitting the private organization of violence, yet keeping it to a self-defensive minimum. Its goal is to restore the democratic order, not to destroy its enemies.125