Bernardo Arévalo’s unexpected second-place finish in the first round of Guatemala’s June 25 presidential election marked one of the most remarkable upsets in Latin America’s recent electoral history, shocking even many of his most faithful supporters. At a Semilla rally the next day, one longtime party organizer ventured a single-word explanation of how Arévalo had managed to sneak into the runoff after polls had shown him trailing in eighth place just days earlier: “Magic,” the organizer told us without skipping a beat. “There’s no other explanation.”
It is easy to understand why so many Guatemalans were stunned. Not only had the Semilla candidate languished in the polls, but the country’s political elites had used almost every tool at their disposal to engineer a runoff between frontrunner Sandra Torres and either Edmond Mulet or Zury Ríos, candidates closely linked to the establishment. For weeks, a Torres-Mulet or Torres-Ríos runoff seemed like a foregone conclusion.
From a data perspective, Guatemala’s June 25 election was indeed among the most extraordinary in the region’s recent electoral history. Arévalo more than quadrupled his expected vote share to leapfrog six other candidates. Strikingly, the election also involved the second highest number of presidential candidates and produced the single most blank and null votes in any free Latin American election since 2015. Now, Guatemala’s democratic future hinges on whether an entrenched establishment succeeds in weaponizing the country’s high courts to reverse this historic result.
We compared polling and election outcomes for presidential candidates who obtained five percent or more of the vote in every democratic election in Latin America since 2015, capturing at least one contest in 16 countries involving 133 contenders. In countries with runoff systems, we considered only the first round of voting. Because our analysis assumes that elections are generally free and fair, we included a total of 27 contests, excluding four marked by serious irregularities —Bolivia (2019), Venezuela (2018), and Nicaragua (2016 and 2021)— as well as the 2017 Honduran election, due to lack of reliable polling.
We looked at the closest survey to each election day that included all relevant candidates. When two or more were conducted in the same period, we took the average of the results. If serious doubts arose about the survey’s validity —usually because the results differed drastically from others around the same time as well as from election outcomes— we took the average of the last and second-to-last polls.
Arévalo outperformed the polls by 8.9 percentage points, jumping from an expected 2.9 percent to 11.8 of the vote — 4.06 times larger than expected. Only one candidate was a bigger overachiever on this metric: two-time Guatemalan contender Manuel Villacorta, who in 2019 won 4.35 times as many votes as expected. But Villacorta finished only seventh, with 5.22 percent. Xavier Hervas, who won 4.0 times as many votes as expected in Ecuador in 2021, finished fourth, albeit with 15.68 percent.
Only nine candidates outperformed by more raw points than Arévalo. One of these overachievers, President Luis Arce of Bolivia in 2020 benefited from candidates dropping out shortly before election day. Another, President Danilo Medina in the Dominican Republic, won the 2016 election with the opposition crying foul play. Former Costa Rican President Carlos Alvarado in Costa Rica participated in a 2018 contest marked by accusations of poll tampering.
And only two candidates jumped forward as many places: Villacorta, who in 2019 was projected in thirteenth, and former Peruvian President Pedro Castillo, who won his short-lived presidency after jumping from seventh to first place in 2021. Of the 27 elections surveyed, 21 involved seven or more presidential candidates.
But Arévalo was not the night’s biggest winner. Neither was Torres, who secured more votes than any other candidate with 15.9 percent. The gold medal instead went to null and blank votes: 24.4 percent —a startling one in four of those who attended the polls— chose not to vote for any of the 22 presidential candidates. Roughly four in ten eligible voters stayed home.
This is the highest percentage of null and blank votes in any of the 27 elections we examined. Peru’s 2021 and 2016 elections had the second and third highest number of null and blank votes, with 18.7 and 18.1 percent, respectively. Guatemala’s 2019 election came in fourth place with 13.1. Across all 27 elections, the average percentage of null and blank votes was 6.2 percent, meaning Guatemala’s first round this year registered four times the average.
Many factors combined to bring about Guatemala’s extraordinary election. The country’s highly fragmented field of presidential candidates surely played a role. A total of 22 candidates made the ballot. Only Costa Rica’s 2022 election included more contenders, with 25. But in Guatemala, had three controversially excluded candidates —Roberto Arzú, Thelma Cabrera, and Carlos Pineda— been allowed to run, the June contest would have tied the record.
Arévalo’s unexpected performance caught the country’s political elites off guard. Alleging fraud, many of them are now leaning on co-opted state institutions, like the Constitutional Court (CC) and the Supreme Electoral Tribunal (TSE), to protect establishment interests and fend off a Torres-Arévalo runoff.
Our results confirm that Arévalo’s display was remarkable even by regional standards. But in no way do they support claims that the June 25 election was fraudulent. We know of no credible evidence to support those allegations.
On the contrary, our analysis highlights the severity of the Guatemalan establishment’s ploy to challenge the results. Guatemalans cast a historic number of blank and null votes and lifted a reformist candidate to one of the most notable upsets in the recent history of Latin America. This is about as clear an anti-establishment message as Guatemalans could have delivered in a process that had been stacked to favor the country’s ruling elites.
If these same elites now succeed in reversing the outcome of the election, Guatemala’s beleaguered democracy could be pushed past the point of no return.
Manuel Meléndez-Sánchez, a PhD candidate in Political Science at Harvard University, is a USIP-Minerva Peace and Security Scholar at the U.S. Institute of Peace. Lucas Perelló is an Assistant Professor of Political Science at Marist College.