The Next Liberal Order
The Age of Contagion Demands More Internationalism, Not Less
With Libya still in the hands of armed regional and tribal factions -- each challenging the other's pretensions to political authority -- it seems wishful to believe that the country will enjoy a smooth and quick transition to stable democracy. Even so, Libya's National Transitional Council and the United Nations are already planning for Libya's first elections.
Soon after the NTC won control of Tripoli, Mustafa Abdel-Jalil, the council's chairman, called for a new constitution and elections within 18 months. An internal UN document, meanwhile, envisions a two-stage transition to democracy in Libya. The first would be a loosely specified period of time during which "political preconditions" for elections -- establishing public security, building public trust in the impartiality of police, and electing a Provisional National Council within six to nine months to write a constitution -- would be satisfied. That would be followed by a six-month period during which the NTC would set up Libya's new electoral machinery, with help from the United Nations.
The UN memo is right to stress the need for preconditions. Our research on all first elections after civil wars since 1945 underscores the dangers of hasty voting. We found that the sooner a country went to the polls the more likely it was to relapse into war. On average, waiting five years before holding the first election reduced the chance of war by one-third.
This makes sense. After civil wars, the rule of law is weak. In addition, those contending for power are usually the same individuals who were recently fighting. The factions that form around them are generally based on traditional social groupings, such as tribes, ethnic groups, and religious sects. In such a situation, candidates resort to illiberal populist appeals, especially ones based on exclusive group identity. Their supporters often refuse to accept election results peacefully, which is especially dangerous if the factions are not yet disarmed and demobilized.
For democracy to take hold, a country needs parties and civic organizations that bridge traditional divides. And we found that even partial demobilization before elections reduces the chance of renewed fighting significantly. But four decades of Muammar al-Qaddafi's rule have left Libya without a civil society, and the NTC will not be able to build one instantly. Moreover, Libya is still awash in weapons, including stocks looted from government warehouses. Those arms are held by rival factions and private citizens alike.
There are some conditions that decrease the probability that even an early election will end in violence. First, if one side is completely defeated, the chance that the election will provoke renewed fighting is cut in half. This sword cuts both ways in Libya: the rebels won a decisive victory over Qaddafi, but the balance of power among the victorious factions remains in flux. The presence of robust international peacekeeping forces during early elections in past cases has dampened the risk of renewed fighting by about 60 percent, all other things being equal. But no one imagines that UN peacekeepers will play a significant role on the ground in Libya.
Second, post-electoral violence is significantly less likely when the country has had a chance to build up impartial, rule-based, and non-corrupt institutions, including courts, police, and other governmental bureaucracies. It is generally better to wait to hold elections until administrative institutions are strengthened, as measured by the bureaucracy's level of expertise, its autonomy from political pressure, and the professionalization of recruitment and training methods. Qaddafi left Libya bereft of modern institutions, and its oil-based economy has allowed for rampant corruption. Overcoming these electoral risk factors will take time.
Finally, agreements among factions to share power also reduce the risk that voting will lead to violence. With such agreements in place, factions have less to lose in an election, so they are less likely to reject the results and return to war. One way to assess whether power in a country is sufficiently shared is to measure the extent of political decentralization, including regional autonomy arrangements. In one test, we found that regional decentralization reduced the likelihood of renewed warfare by four-fifths. As of yet, Libya has no such agreements.
All this means that the NTC proposal to hold elections within 18 months is imprudent. Meanwhile, the desiderata and preconditions mentioned in the UN memo range from glittering generalities about political inclusiveness and human rights to the procedural specifics of setting up the constitution-making process. The vital political part of the plan is vague, and the more concrete legalistic part begs the question of whether political realities will allow that process, even if built, to run smoothly. It will not be useful if legal formalities are pushed ahead of political realities.
In the past, the international community has often contributed to the quick election problem, first by pressing warring factions to reach precarious settlements before either side has won decisively, and then by urging fast elections. In the particularly tragic case of Burundi, international donors demanded that the ethnic minority Tutsi military regime hold elections in 1993 following its armed repression of a series of regional Hutu rebellions. This led to the election of a Hutu president, his assassination by the military, and an ethnic bloodbath that killed over 200,000 people.
In principle, international actors can mitigate the problems of early elections if they provide a robust peacekeeping force, facilitate the demobilization of armed groups, support reliable power sharing agreements, and help to build modern political institutions. In Libya, the size of these tasks is great because of the country's vast institutional deficit, the multiplicity of its armed factions, and its corruption. Moreover, international involvement is likely to be limited. All in all, our research provides a warning that early elections under such conditions will increase the risk of a return to civil war.