A steady drip of claims that next Tuesday’s election is at risk of rigging, and explicit calls by some far-right actors to their supporters to show up at polling places, have raised the specter of violence on Election Day or thereafter. “Go around and watch other polling places,” Donald Trump has said repeatedly. Neo-Nazi and other extremist groups have been blunt: The Daily Stormer, which bills itself as “the most visited alt-right website,” advised its readers that “if we don’t win this, it’s going to be open race war” alongside illustrations of Hitler and guns. And researcher Jonathon Morgan found an exponential spike in activity on some 246 Facebook pages maintained by U.S. militia groups, with explicit posts about the possibility of post-election violence.

Violence was once a regular feature of American elections. But because America has been blessed with electoral peace for several decades—and perhaps because white Americans in particular have experienced very little electoral discord in the last century—American commentators have treated 2016 as if it were entirely sui generis, and as if we have few or no tools to think about the specter of societal unrest, or how we might prevent violence and promote healing.

But that is wrong. Globally, we do have a body of knowledge that can help indicate when polarizing electoral seasons are likely to produce violence and unraveling. Even better, we have a set of ideas on how to empower citizens and institutions to defend themselves. And best of all? Though these practices are now known primarily in the development and peacebuilding communities, many originated in polarized American communities a generation ago.

Across the globe, electoral violence is widespread and diverse. In 2014, 27 countries on five continents experienced it. And this ranged from India—a vast country with well-established democratic institutions, which nonetheless has violence as a recurring feature of voting and civic life—to South Africa, where electoral violence has steadily decreased in magnitude in the years since the establishment of full democracy in 1994; from wealthy Bahrain to conflict-torn Afghanistan.

Because conflict prevention and resolution experts have long identified elections as a potential flashpoint for violence, the literature on the connection between the two is vast. Practitioners in affected countries, international organizations, and Washington’s own U.S. Agency for International Development (USAID) have developed extensive frameworks for predicting, preventing, and preparing to manage the reality of election-season violence, and to heal in its aftermath. This literature is a thought-provoking and sobering read for Americans contemplating the future of our politics after this cycle, starting with a U.S. Institute for Peace (USIP) assessment essay here and a helpful synthesis of three influential approaches here.

Don’t have time to catch up on all the literature before Tuesday? Not to worry: Below, I offer four of the most important takeaways—innovations and experiences that can make a real difference in how states and municipalities, law enforcement and electoral agencies, as well as public figures and ordinary citizens can prepare for the upcoming election, make sense of what happens, and play constructive roles in the aftermath.

See the Picture Beyond the Campaign, or: It’s Not (all) Donald Trump’s Fault

Election-specific issues, from hate speech to fraud allegations to faulty election procedures, are important, but are not by themselves sufficient to trigger or predict violence. Some, like the proximity of war or insurgency, authoritarian or single-party rule, are blessedly inapplicable to the United States.

And others are obvious and have indeed been topics of conversation in our election: namely income inequality, as well as ethnic, racial, and religious strife.

Others, though well-documented, may be surprising to Americans. High rates of gender discrimination and gender-based violence are strongly associated with the risk of communal conflict. First-past-the-post electoral systems are believed to be more vulnerable to violence because they create winner-take-all dynamics. And yes, places with the easy availability of guns and the significant presence of private militias are more likely to experience election violence than those without them.

Seasoned election observers (and, quietly, our own law enforcement agencies) are well aware that these factors differ enormously across various regions of the United States. Election violence theory also tells us that a history of conflict matters; applying this theory to the United States would point you toward areas most affected by violence from the Civil War and Reconstruction, the civil rights movement, as well as more recent violence from nationalist groups of all kinds.

Random Individuals Don’t Get “Caught Up” in Violence

Reporting on fears of such violence in the United States paint a picture of rowdy mobs and chaos. Instead, the research literature tells us to expect electoral violence to be instigated by political actors, with mobs and chaos arising in response to pre-planned provocations and acts of violence. In sub-Saharan Africa, election violence is often “orchestrated by elites to intimidate voters.” Yale’s Steven Rosenzweig, while casting doubt on the assumption that hate speech and violence are efficient means to affect voters’ choices, found that political elites believe either that they are completely irrelevant or that they can indeed sway voters. Most concerningly, he also found that hate speech in campaigns does increase the risk of violence.

When Ghanaians were surveyed about violent actors in their local elections, they showed clear understanding of this elites-to-instigators-to-masses dynamic by pointing their fingers at the country’s youth: because they “allow themselves to be manipulated by politicians,” party supporters and functionaries, and so-called “thugs.” Violence prevention and de-escalation strategies thus have three audiences: Political elites, the public at large, and individuals and communities deemed most likely to be pressured to participate in such violence.

Technology Can Escalate or De-escalate

We’re also used to hand-wringing about how new media technologies may make violence more likely. They allow us to create cocoons of incorrect and inflammatory information, minimize time for real reflection, and are used to organize mass gatherings which may not be peaceful. The international casebook, though, tends to focus more on the role information technology and social media—in the hands of citizens, no less—play in spotting violence quickly and in connecting forces with the power to protect or defuse.

Open-source geographical information systems (GIS) platforms have been developed that allow users to report violence, threats, and election irregularities via text message or social media, and then map them to track danger and plan a response. One, Ushahidi.com, has been used in Mozambique and Kenya to empower civilians to monitor elections and report irregularities or violence. Lest it sound like this is a tool for developing countries only, different applications of this platform were used in 2012 by both the Obama campaign to get voters to the polls in key areas, and by a coalition of civil rights and legal advocacy groups, where it largely responded to concerns of voters displaced by Hurricane Sandy.

Another combines information technology with cultural expectations around the role of women to produce powerful responses to violence and intimidation. Threats of violence against women are often an explicit component of different strategies to deter voters or punish voting through violence. Analysts identified a spectrum, which, at its extreme, ended in rapes, killings, and threats aimed at Kenyan women during the country’s 2007 election, but which begins with the kind of online abuse and public taunting of women that has marked the 2016 campaign in the United States. In response, a form of community-level election monitoring called Women’s Situation Rooms recruits and trains women and young people as community monitors. Their cell phones and text messages are fed to a network of analysts, also connected to a team of eminent persons, both domestic and international, who can be called upon as mediators if the efforts of the women or local law enforcement are unsuccessful. This was developed by Liberians and has been deployed in Kenya, Senegal, and elsewhere to considerable praise. One could imagine community-level coalitions of religious leaders playing similar roles in some contested parts of the United States.

Training’s Value—and its Limitations

India, which suffers from chronic patterns of electoral violence, has won praise for its ability to cross-train law enforcement and electoral officials to manage and contain conflict as it arises. The Women’s Situation Room, and other African models, places trained mediation teams on call at polling sites. Experts stress the need to begin planning, training, and team-building as much as eighteen months before elections take place. And it is hard to blame U.S. officials for failing to anticipate, when a Jeb Bush versus Hillary Clinton campaign seemed a safe bet, that polling-place mediation might be necessary. But both parties have extensive networks of lawyers available on Election Day; one can imagine national networks of mediators making their services available on standby to local electoral authorities, especially where trained mediators have their roots in local communities. One could also imagine a few conference calls between law enforcement, election boards, and key U.S. repositories of information and experience in election violence prevention like the Alliance for Peacebuilding and the U.S. Institute for Peace.

However, USIP’s Jonas Claes sounds a cautionary note about training and programming against election violence, which is, perhaps, the most important takeaway for Americans this year. Six years after presidential elections in Kenya in which more than 1,000 died, interviewees described the results of intensive domestic and international interventions—which did help ensure largely peaceful follow-on elections—as “negative peace,” in which conflict was suppressed rather than resolved. “Palpable tension, fear and anxiety” were the result.

Whether November 8 brings violence to the United States, international experience would warn us that the risk factors for political violence won’t simply go away when the campaign attack ads come off the air. Americans are used to thinking of political violence as something that happens somewhere across an ocean, in unfamiliar surroundings. But this year, with unnerving and violent parallels to our own politics from the Philippines to Central Europe, it wouldn’t hurt us to prepare.