As policymakers and stakeholders grapple with designing post-war strategies in Iraq, Libya, Syria and Yemen, it is increasingly important to understand what works to prevent countries from relapsing into war.

Conflict-torn states have narrow windows of opportunity to prevent war relapse, and failure is common. About half of all post-war countries lapse back into civil conflict between the same belligerents in the first decade after the end of fighting.

In a recent paper for the United Nations in Beirut, I analyzed all 109 post-war episodes since 1970 to identify risk factors for war relapse. The paper drew on the large quantitative literature on civil war and post-war recovery, including work from the Uppsala Conflict Data Program, the Peace Research Institute Oslo, the Polity IV project and the World Bank.

Long-standing international conventional wisdom prioritizes economic reforms, transitional justice mechanisms or institutional continuity in post-war settings. However, my statistical analyses found that political institutions and military factors were actually the primary drivers of post-war risk. In particular, post-war states with more representative and competitive political systems as well as larger armed forces were better able to avoid war relapse.

These findings challenge a growing reluctance to consider early elections and political liberalization as critical steps for reestablishing authoritative, legitimate and sustainable political order after major armed conflict.

The non-results are perhaps as interesting as the results. With one exception discussed below, there is no evidence that the economic characteristics of post-war countries strongly influence the likelihood they will return to war. Income per capita, development assistance per capita, oil rents as a percent of GDP, overall unemployment rates and youth unemployment rates are not associated with civil war relapse.

Equally significant is there is no evidence that the culture, religion or geopolitics of the Middle East and North Africa will impede post-war recovery. I introduced into the statistical models measures for Islam, Arab culture and location in the region. None of these variables showed statistically significant correlations with the risk of war relapse since 1970, holding everything else constant, suggesting that such factors should not distinctively handicap post-war stabilization, recovery and transition in Iraq, Libya, Syria or Yemen.

What did matter, then? I found three primary sets of drivers of post-war relapse into conflict.

1. Political institutions of post-war countries

Democracies have an 82 percent lower risk of relapse relative to all other regime types. Having held nationwide elections within the previous five years reduces risk of war relapse by about 56 percent. Delaying elections is a risky strategy: The likelihood of relapse increases six percent for every year that post-war elections are postponed. The results suggest that political competition and democratic legitimacy can greatly reduce risks in post-war settings.

Decentralization also affects post-war risks, but differently as time passes. Decentralization, including federalism, correlates with lower risk of relapse for roughly the first five years of a peace period, but it correlates with heightened risk thereafter. This finding suggests that decentralization can be an effective way to buy some peace in the immediate aftermath of armed conflict, but in the long run decentralization often sows the seeds of renewed conflict.

One possible interpretation is that decentralization inhibits the ability of central governments to deliver public goods to peripheral regions effectively. Another, somewhat related interpretation is that, in practice, post-war decentralization locks in place local wartime leaders who simply use the peace period to prepare for renewed conflict.

All told, the design of political institutions matters a great deal for shaping post-war risks.

2. Military and security-related factors

Perhaps not surprisingly, larger security sectors reduce the risk of war relapse. For every additional soldier in the national armed forces per 1,000 people, the risk of relapse is about seven percent lower. Larger militaries are better able to deter renewed rebel activity, as well as prevent or reduce other forms of conflict such as terrorism, organized crime and communal violence.

The presence of outside troops also has significant influence on risk. The analysis lends support to a well-established finding in the political science literature that the presence of United Nations peacekeepers lowers the risk of conflict relapse. However, the presence of non-U.N. foreign troops almost triples the risk of relapsing back into civil war. There are at least two potential interpretations on this latter finding: Foreign troops may intervene in especially difficult circumstances, and therefore their presence indicates the post-war episodes most likely to fail; or foreign troops, particularly occupying armies, generate their own conflict risk.

3. Economic growth

The only economic variable found to be associated with post-war risk is income growth per capita. Notably, like decentralization, the relationship with risk changes over the course of the post-war period. In the initial post-war months, a five percent annual GDP growth rate per capita is associated with 25 percent greater risk of relapse, relative to flat growth. Over time, this economic growth-related risk decreases, until about 20 months into the peace period, when higher growth becomes a protective factor and begins mitigating the risk of war relapse. By month 75, a five percent annual growth rate is associated with a 50 percent reduction in risk of relapse, relative to flat growth.

This finding suggests that economic stabilization and recovery are not quick-win strategies to prevent war relapse, contrary to conventional wisdom in the international community. In the initial post-war years, high growth is associated with higher risk of war relapse, and the pacifying effects of income growth only kick in after two years. In the short run, rapid growth cannot substitute for the careful design of transitional political and military frameworks and may even prove destabilizing. In the long run, however, countries can see dramatic reductions in their risk of war relapse by maintaining high growth rates.

These economic gains would only be realized, though, through the careful design of the political institutions and military architectures of post-war societies.

George Frederick Willcoxon is a political scientist working at the United Nations. The opinions expressed herein are those of the author and do not necessarily reflect those of the U.N. Follow him @georgewillcoxon.