Diplomats and international election experts have worked with police, supported election commissioners or pressured lead candidates to refrain from inciting violence. Civil society groups often focus on involving youth in constructive activities and using peace messaging on social media or in other forums. But what works in a given context, what doesn’t and how can assistance programs be more effective?

This Q&A is based on an interview the European Institute of Peace conducted with Claes, who recently completed a year in residence at the Brussels-based organization. His book was the subject of a symposium of European Union officials, members of the European Parliament and other experts at the European institute on Nov. 16. The event was organized with USIP, the European Parliament’s Democracy Support and Election Coordination Group, the European External Action Service and the European Peacebuilding Liaison Office.

Symposium participants largely agreed on the need for more international engagement between elections. They also said security assessments must be conducted much earlier and involve a broader range of participants. Follow-up to election-observation missions needs to be stronger, too, taking advantage of useful opportunities to work with new administrations to address flaws that became evident during the election process and to carry out much-needed reform.

How common is election violence? How big of a problem is it?

Election violence presents a major problem. According to the African Electoral Violence Database, nearly half of the countries in sub-Saharan Africa regularly experience some form of election violence. Failed prevention may require costly interventions, reverse years of development and generate a deep distrust in democratic governance.

But the scope of the challenge should not be dramatized either. Even in sub-Saharan Africa, only 10 percent of elections involve the type of widespread killings we saw in Kenya in 2008, or in Nigeria in 2011. On average, election years are also no more violent than other years. But more so than other types of political violence, election violence undermines the legitimacy of political leaders and a country’s regard for democratization.

Could you briefly summarize the main findings of your book?

Sure. Together with an impressive team of researchers, I closely studied several elections—Bangladesh, Moldova, Honduras, Thailand, Malawi—that were roughly at equal risk of election violence. In the end, each election experienced very different levels of violence. This set-up allowed us to analyze whether prevention made a difference.

Based on our findings there are strong indications that prevention works, but each approach is not equally promising. Those efforts by the state apparatus in guaranteeing election security and administrating its elections, through its police units or election commission, seemed the most relevant. The impact of engaging local youth or distributing peace messages was less pronounced.

This is quite surprising, no? Youth engagement and peace messaging seem to be quite popular approaches.

That is right.  Both techniques are frequently used by local civil society, as they target a broad section of the electorate. Peace messages encourage voters, parties, and candidates to refrain from violence through sports events, cultural activities or social media. Youth education, employment or training programs try to engage young people specifically. Both approaches aim to change attitudes and behaviors in a very short timeline, but often fail to realize these ambitious objectives. Another common civil society approach, civic education, seems more promising.

You indicate that state institutions hold the key to peaceful elections. Does that mean international efforts are a waste of time and resources?

Certainly not. Surely the state has a primary responsibility to organize the elections and guarantee security—and their efforts seem to matter the most. But neighboring countries, regional donors and international civil society can play an important secondary role in building the capacity of local police, the election commission or a country’s judicial system. The monitoring missions of international organizations can indirectly reduce the risk of violence as well, by channeling critical intelligence, deterring potential culprits or by validating the quality of the electoral process.

At times, however, the international community overestimates its own impact. This dynamic was very clear during the 2013 elections in Kenya. The vote was relatively calm, and did not see a widely anticipated repeat of the mass killings that occurred in 2008. However, our research found that the relative calm did not result from international prevention, but rather from the fear and memory of past inter-ethnic clashes around the elections. Having lost faith in political leaders and hope in the credibility of the electoral process, many Kenyans also voiced their unwillingness to fight the results.

The European Institute of Peace is actively engaged in political dialogue. How can dialogue serve to prevent election violence most effectively?

Dialogue is quite prevalent at the community level. One promising model that USIP applies is the engagement of citizens and police in dialogue about justice and security priorities. These initiatives may improve community-police relationships, and identify common security priorities. In Liberia, local efforts are also underway to connect the youth representatives of political parties with police officers through dialogue.

The European Institute of Peace seems primarily engaged in direct and high-level political dialogue. The impact of this approach is promising, but merits further research. In fact, preventive diplomacy stood out in our research as the only instrument that corresponds with higher violence levels. Evidently there is no causal effect at play here—preventive diplomacy does not cause more violence. But the relationship confirms how diplomacy is often used as a crisis management tool, when violence is imminent or already ongoing.

Special envoys, cables or public statements are usually issued late in the game, as a last resort, and not in a truly preventive sense. USIP is searching for ways to make diplomacy truly preventive, considering those approaches that can change the decision-making calculus of leading politicians and candidates early on in the election cycle. The Institute’s evaluative research currently continues in Kenya and Liberia, where we are closely reviewing prevention efforts taken prior to the 2017 elections, to identify those approaches that seem to be the most effective.

Wouldn’t the success of prevention depend on the context of a specific election?

The success of prevention is indeed highly contextual—but it is important to engage strategically, based on a thorough assessment, and to start early. When an election date is set, and the type of risk is identified, carefully selected prevention efforts should start immediately. Too often, we apply the same measures to different situations, or we wait until three or four months before election day to initiate programming. The window of opportunity is long gone by that time.