The ten-point lead had evaporated in seven days
A week before Peru's June 6 presidential runoff, a nation long divided by geography, class, and memory finds itself at an electoral crossroads: the rural schoolteacher Pedro Castillo and the political heir Keiko Fujimori have converged into a statistical tie, their fortunes shifting four and five points in a single week. What the numbers reveal is not merely a horse race but a map of Peru itself — its interior against its capital, its poor against its prosperous, its hope for change against its fear of it. With millions still undecided or planning to reject both candidates, the country's next chapter remains unwritten.
- A ten-point lead that felt like destiny dissolved in seven days, turning Castillo's commanding advantage into a margin too thin to call.
- The poll lays bare a country split along fault lines older than any campaign: Lima votes one way, the Andean south another, and the two Perus are barely speaking the same electoral language.
- Both candidates carry the weight of mass rejection — 43% of voters refuse Fujimori, 40% refuse Castillo — making this less a contest of enthusiasm than a referendum on which risk the country is willing to take.
- Fujimori is consolidating Lima and upper-income voters while Castillo bleeds support, suggesting the capital's gravitational pull may yet reshape the outcome.
- With 6.3% undecided and 13% poised to cast blank or null ballots, the final week is less about persuasion than about which campaign can convert reluctance into a vote before June 6.
Seven days before Peru's presidential runoff, what had looked like a decisive lead had become a coin toss. An IEP poll conducted for La República placed Pedro Castillo, the leftist schoolteacher from Cajamarca, at 40.3 percent — and Keiko Fujimori, daughter of the imprisoned former president, at 38.3 percent. Within the poll's 2.8-point margin of error, the two were effectively tied.
Just a week earlier, Castillo had held a ten-point advantage. Between May 23 and the new survey, he shed 4.5 points while Fujimori gained nearly four — a dramatic reversal with less than a week until voters decided. The poll reached 1,227 respondents across 24 departments by telephone on May 27 and 28.
The data sketched a portrait of a divided country. Castillo dominated rural Peru with 51.5 percent to Fujimori's 27.1, and held leads across the south, center, and east. But Lima told a different story: Fujimori commanded 51.1 percent in the capital while Castillo managed only 27 — a 24-point deficit in Peru's most populous region. The north was the one place they were statistically even.
Class structured the race just as sharply as geography. Lower-income voters in the D and E brackets favored Castillo 45.9 to 30 percent. Wealthier Peruvians in the A and B brackets backed Fujimori 53.1 to 34.5, and the middle class leaned her way as well.
Beneath the horse-race numbers lay a deeper unease. Fujimori's voter rejection stood at 43 percent — an improvement from 50 percent the week before — while Castillo's had climbed to 40 percent. Another 13 percent planned to cast blank or null ballots, and 6.3 percent remained undecided. With both candidates carrying heavy negative ratings and momentum visibly shifting, the outcome of June 6 was, by any honest measure, genuinely open.
Seven days before Peru's presidential runoff, the race had tightened into a statistical dead heat. An Instituto de Estudios Peruanos poll conducted for La República showed Pedro Castillo, the schoolteacher from Cajamarca running on a leftist platform, at 40.3 percent support. Keiko Fujimori, the opposition leader and daughter of former president Alberto Fujimori, stood at 38.3 percent. The margin of error was 2.8 percentage points, meaning the two candidates were effectively tied.
Just one week earlier, Castillo had held a commanding position. The previous IEP survey, published on May 23, showed him with 44.8 percent to Fujimori's 34.4 percent—a ten-point advantage that had evaporated in seven days. Castillo lost 4.5 points while Fujimori gained nearly four, a dramatic swing in a race that would be decided on June 6. The poll was conducted by telephone between May 27 and 28, reaching 1,227 respondents across 24 departments.
The numbers revealed a country divided along geographic and economic lines. Castillo dominated rural Peru, where he claimed 51.5 percent support compared to Fujimori's 27.1 percent. In urban areas, his lead narrowed but held: 45.3 percent to her 33.7 percent. But in Lima Metropolitana, the capital and its suburbs, the picture inverted completely. Fujimori commanded 51.1 percent there while Castillo managed only 27 percent—a 24-point deficit in the nation's most populous region.
Regionally, Castillo's strength lay in Peru's interior. He led the central region with 49.5 percent versus Fujimori's 28.9 percent. In the south, his advantage was even starker: 56.3 percent to 19.6 percent. The eastern regions showed similar patterns, with Castillo at 48.6 percent and Fujimori at 30.3 percent. The north was the only region where the two were statistically tied, with Fujimori at 41.7 percent and Castillo at 39.4 percent.
Socioeconomic class also structured the vote. Castillo held the support of lower-income Peruvians in the D and E brackets, where he polled at 45.9 percent compared to Fujimori's 30 percent. Fujimori dominated among wealthier voters. In the A and B income brackets, she reached 53.1 percent while Castillo managed 34.5 percent. The middle class, classified as C level, favored her as well: 42.3 percent to 34.8 percent.
But the poll also captured something else: the depth of rejection both candidates faced. Fujimori's antivoto—the share of voters who said they would not support her—stood at 43 percent, though this represented improvement from the previous week's 50 percent. Castillo's rejection had grown to 40 percent, up from 37 percent seven days earlier. Additionally, 6.3 percent of respondents remained undecided, while 13 percent said they would cast blank or null votes, further complicating the final week's dynamics.
The narrowing of the race suggested momentum had shifted. Castillo's initial lead, built on rural support and leftist sentiment in Peru's interior, was eroding as the runoff approached. Fujimori, despite her higher rejection rates, was consolidating support among Lima's voters and the country's wealthier segments. With less than a week remaining and both candidates carrying substantial negative ratings, the election remained genuinely uncertain.
Notable Quotes
Castillo's support fell from 44.8% to 40.3% in one week while Fujimori rose from 34.4% to 38.3%— Instituto de Estudios Peruanos poll for La República
The Hearth Conversation Another angle on the story
How did Castillo lose four and a half points in a single week?
The poll doesn't explain the mechanism, but you can infer it. He'd been the frontrunner, which means scrutiny intensifies. Fujimori's campaign had a week to define him to undecided voters, especially in Lima. And some of his supporters may have gotten nervous—they saw the numbers tightening and wondered if he could actually win.
Why does Lima matter so much if he dominates the countryside?
Because Lima is where the money is, where media is concentrated, and where roughly a quarter of Peru's voters live. You can win the south and the highlands decisively, but if you lose the capital by 24 points, you're fighting an uphill battle in a national race.
The rejection numbers are striking—43 percent won't vote for Fujimori, 40 percent won't vote for Castillo. What does that mean for turnout?
It means both candidates are polarizing. But it also means there's a real pool of swing voters—the 6.3 percent undecided, plus people who might change their minds about voting blank. In a technical tie, those voters become everything.
Did Fujimori's antivoto actually improve, or is that just noise?
It improved—she dropped seven points in rejection in one week. That's real movement. She was at 50 percent rejection before; now she's at 43. She was successfully rehabilitating her image in the final stretch, at least among some voters.
What about the blank and null votes—13 percent is substantial.
It's a protest vote, mostly. Voters who dislike both candidates enough to spoil their ballot rather than choose. In a dead heat, that's a real factor. If even half of them break toward one candidate, it could decide the election.