By Clayton Besaw THE insurrection at the US Capitol on January 6 shocked Americans and the world. But the US was not alone in its rocky transfer of power: Last year saw more election-related violence than any year in the past four decades. At least 44% of national elections that took place in 2020 had some form of violence, according to a new election violence database my team of conflict analysts released in 2021. That is 33 national votes marred by violence. Election violence is a type of political violence that seeks to unduly influence the process and outcome of a vote. Attacks can be perpetrated by governments or civilians, against electoral infrastructure, political parties or voters. Our database tracks all incidents of election violence going back to 1975. Until last year, the worst year was 1990, when violence occurred in around 46% of elections. The median annual rate dating back to 1975 is 30%. How will 2021 stack up? Countries to watch So far, this year looks to be slightly more peaceful than last. Of 42 national elections scheduled to occur in 2021, our data forecasts that roughly 40% are at high risk of violence. Chad, Ethiopia and Haiti are at the most risk for election-related violence this year. Chile, a generally stable democracy, is also unusually high on the list after a year of civil unrest. Why these countries? Though the causes of election violence are complex and unique to each country, my research finds many places that experience election violence share the same five characteristics: a history of election violence (indicating violence is “normalised”), high coup risk (an indicator of a weak State), high infant mortality rate (an indicator of a weak State), pervasive political violence (an indicator of a weak State), low gross domestic product (indicating a struggling economy). When all five characteristics are present, our early warning forecaster raises red flags. How we predict violence The national data we processed finds that countries with longer and more recent histories of election violence are at risk of more violence. The presence of other kinds of political violence — terrorism or civil war, for example – also increases the risk of election violence because political attacks can come to be seen as “normal.” The government’s ability to exercise power is another important factor in election violence risk. That’s because authorities in weak States usually cannot physically stop clashes — even if, as was the case in Kenya’s 2007 election, they can reasonably foresee conflict erupting. History shows a weak government may also inflict election violence if the incumbent fears losing political power. That’s what Azerbaijan’s president did in 2005 when growing political opposition threatened his leadership before the election. To measure State capacity, we consider infant mortality rate, gross domestic product and the risk of regime change. These factors show the government’s ability to maintain stability and promote well-being, which are reliable indicators of its power. Election violence