Women line up outside a polling station in Herat, Afghanistan, during parliamentary elections Saturday.  (Hamed Sarfarazi/AP)

The Taliban promised to disrupt Afghanistan’s national elections over the weekend and did just that, launching deadly attacks on polling stations. In Africa, a report tallies more than 380,000 dead in South Sudan’s civil war — as the country’s president and rebel leaders reach a new peace deal.

Every day, these types of reports tell a familiar story: There was violence, the United Nations got involved, donor countries pledged millions in assistance, warring parties signed agreements, and headlines praised peace. But the violence continues.

We tend to look at what went wrong when we’ve tried to resolve conflicts. My new research takes a different approach and looks at what has gone right.

To stop wars, an essential first step is to understand how people have actually succeeded in building peace.

Most politicians assume that it takes an agreement between world elites to end armed conflicts. So they organize conferences to reconcile governments and rebel leaders, they push for general elections, and they strive to reconstruct state bureaucracies: This was the plan in Afghanistan, Congo, Iraq, South Sudan and so on.

But top-down peace isn’t working

In short, leaders try to build peace from the top down. They also rely on the knowledge and resources of diplomats, U.N. peacekeepers and other foreign interveners.

This approach has had disastrous consequences. Battle deaths have risen by 340 percent globally in the past 10 years. The world has 1.5 billion people living under the threat of violence in more than 50 conflict zones. Wars have recently triggered the worst refugee crisis of the past 70 years.

Here’s an exception — Idjwi, a Congolese island of peace in Lake Kivu. For the past 20 years, one of the deadliest conflicts since World War II has raged in Congo. Even the largest U.N. peacekeeping mission in the world has failed to stem the flood of violence. Idjwi somehow avoided the conflict, although the island contains all of the ingredients that have caused generalized fighting in other parts of Congo: geostrategic location, mineral resources, ethnic tensions, lack of state authority, extreme poverty, and conflict over land and power, to name a few.

How to promote a culture of peace?

Idjwi is peaceful thanks to the everyday involvement of all of its residents, including the poorest and least powerful. The local population has fostered what they call a “culture of peace.” Local associations help defuse tensions by mediating between parties in conflict. Priests often act as peacemakers. Blood pacts (traditional promises between two parties who agree never to hurt each other) link most of the island’s families. And Idjwi’s reputation as the “land of the sorcerers” scares away potential troublemakers from neighboring provinces.

There are other cases of peace amid extreme violence, from Somaliland (north of Somalia) to various Colombian villages. As in Idjwi, it is not the army, state or police — or any outside peacekeeper — that keeps tensions in these areas from erupting into violence but the community members themselves.

Of course, civilians cannot defeat armed groups single-handedly. Nor do ordinary people have the networks necessary to build national peace. Here’s one solution: Elevate local peace-building to a status equal to the top-down strategies we currently use to resolve crises.

But this also means radically rethinking our approach to conflict resolution. Most interveners assume that local people are incompetent or corrupt, even violent, believing that only outsiders have what it takes to build peace.

The stories of Idjwi and other parts of the world show that local people can hold violence at bay. However, they often lack the resources and sometimes the technical skills to implement effective peace initiatives. So they do need foreign support — but not the counterproductive “foreigners-know-best” assistance that we usually see.

There’s another way to build peace

Instead, other peace-building models are gaining traction. The Swedish Life and Peace Institute, for instance, doesn’t assume that it knows what the problems are and how to fix them. In Congo, this nongovernmental organization empowers local people to develop their own analyses of their communities’ conflicts, agree on the best answers and implement those solutions.

Examples like this encourage us to collectively question our assumptions about war and peace. Take, for example, the idea that all good things — like peace, democracy and education — naturally reinforce one another. Although education and electoral projects are the backbone of many peace initiatives, the situations in Afghanistan, Angola, Congo and Rwanda show that prioritizing such ideals without considering the local context can actually end up fueling violence.

What builds peace is not elections. It may not even be democracy — at least not right away. It may not require billions in aid or massive international interventions. And it may involve respecting superstitions and local beliefs.

Ultimately, many successful examples of peace-building in the past few years have involved innovative grass-roots initiatives, led by local people, often using methods the international elite tends to dismiss. But perhaps these very efforts merit a closer look as a first step to changing the way we view and build peace.

Severine Autesserre is a professor of political science at Barnard College, Columbia University. She is the author of Peaceland and The Trouble with the Congo, and she is finishing another book, tentatively titled On the Frontlines of Peace.