In the final votes for the Presidency, the next June 21, Abelardo De la Espriella would defeat Iván Cepedathe candidate supported by President Gustavo Petro, for 7.6 percentage pointsmore than double the advantage he gained in the first round on May 31.
That is the great conclusion of the final survey by Guarumo and Ecoanalíticarevealed by EL TIEMPO, which consulted in person the voting intention of 2,073 Colombians in 54 municipalities and that provides support from 52.6% for Abelardo in front of a 45% of Iván Cepeda. Only 2.4% of those consulted assured that they will vote blank.
According to the measurement, the two candidates who went to the runoff they grew in voting intention with respect to the results of the same survey last May, when the eventual scenario of a second round between the candidates who would actually compete in this instance was asked. Cepeda went from 40% to 45%, while De la Espriella, who surprised in the first round by getting almost 700 thousand votes from the candidate of the Historical Pact, shot up from 43.6% in May to 52.6%, exactly 9 percentage points more. He blank votein turn, plummeted from 16.4% from a month ago to 2.4% of the current measurement, which shows that the polarization that has characterized the campaign continued and will continue to be decisive until the final stretch. Under this logic, white, which today translates into a position and not into undecided people, will not be the protagonist. In fact, it maintains a similar figure to that of the last elections.
He growth in voting intention for De la Espriellain turn, indicates that the electoral strategies deployed by the left-wing candidate and by President Gustavo Petro, who has openly participated in politics despite legal prohibitions, they have not managed to reverse the trend which was already marked in the polls last month and which was ratified at the polls two weeks ago.
Precisely, with respect to the real results of the first round voting, De la Espriella and his formula, José Manuel Restrepo, would grow by 9.2%; while Cepeda and his ‘vice’, Senator Aida Quilcué, would do so in 4.7%.
Víctor Muñoz, political analyst and co-founder of Guarumo, points out that the growth of distances is explained because Abelardo would be receiving the support of most of the 1,600,000 votes that Paloma Valencia achieved on May 31in addition to the 200 thousand votes that businessman Santiago Botero surprisingly hadalso related to the essence of the ‘Tigre’ proposal. He adds that the decision of Sergio Fajardo about not taking sides for no candidate would have also influenced the fact that a good part of the million plus citizens who supported him could have made the decision to go with the opposition candidate. This explains its growth.
“When you look at this 6% of Paloma, practically everything is being added to Abelardo de la Espriella. And the other thing is that the half of Fajardowho got 4%, is also going to that campaign,” says Muñoz about the growth of De la Espriella and says that in the case of the ruling party candidate it adds up “due to the arrival of half of Fajardo’s votes and a bit of the marginality that a Sondra Macollins Garvin, a Roy Barreras and a Claudia López had,” Muñoz asserted.
For analyst Jairo Libreros, these results are a reflection of the strategies that both campaigns adopted to face this second round. It highlights that De la Espriella strengthened the strategies that allowed him to be the most voted on the election day of May 31 and mentions that the Cepeda’s campaign definitely did not shakeeven though there are discomforts both in Casa de Nariño and in the Historical Pact due to how the strategies have been handled.
“The two campaigns that entered the runoff followed diametrically opposite paths. Iván Cepeda’s was characterized by a series of erratic and contentious decisions. On the one hand, ignoring the electoral results of the first round, without analyzing the context of the pre-count, just to avoid conflict with President Gustavo Petro, sent a message of political submission to the House of Nariño and set off all the democratic alarms about its inability to facilitate the peaceful transfer of power based on the popular will expressed at the polls. On the other hand, Abelardo De la Espriella’s campaign did not lose the electoral focus and followed the electoral strategy that was expected for a presidential second round scenario: the effort was concentrated on the conquest of the voting intention of the sectors located in the center of the political spectrum. The figure of his vice-presidential formula starred in marathon days of interventions in the media, on social networks and in academic events with programmatic messages of public policy,” said the academic.
The Guarumo survey, which was carried out between June 8 and 12also consulted the opinion of Colombians on the management of the government of President Gustavo Petro, whose approval grew in the last month. 19% of those surveyed rated it as very good, while 31.3% said it was good. One in four (25.2%) said it was bad, 19.3% considered it very bad and 5.2% did not respond. A striking fact is that the favorable image of Petro, which is close to 50%, does not transfer completely to the candidate for whom he has openly gambled, Cepeda, whose electoral support is five points below.
Guarumo consulted citizens in 54 municipalities who expressed their intention to participate in the second round, using a copy of the presidential card that Colombians will receive at the polls in a week. He overall confidence level is 95% and the overall sampling error margin of 2.9%.
POLICY
Technical sheet
Legal entity that carried out the survey
Guarumo and EcoAnalisis: Measurement and economic concepts
Legal entity entrusted by Guarumo SAS and EcoAnalítico
Measurement and Economic Concepts SAS
Funding source
Guarumo SAS and EcoAnalisis Measurement and Economic Concepts SAS
Purpose of the study To know the voting intention of Colombians for the candidates for the Presidency and Vice Presidency of the Republic of Colombia
In-person survey with probabilistic sampling Elections Presidency of the Republic 2026 Universe represented Set of adults (over 18 years of age) in non-institutionalized homes in Colombian territory (rural and urban), with the intention to vote in the elections that will be held in 2026. The target population has coverage of 99% of the universe under study and includes 21.6 million adults with the intention to vote. The importance or weight of the regions is as follows: Bogotá: 17.7%; Caribbean: 18.6%; Central-east: 20.8%; Central-south: 8.8%; Coffee Axis: 17.9% and Pacific: 16.2%
Sampling frame
The sampling frame that supports the probabilistic design of the sample is the National Geostatistical Framework provided by Dane. In the design of the sample, the population projections for the year 2026 published by Dane and the detailed voting in the electoral contest of May 29, 2022 provided by the National Registry of Civil Status were taken into account.
Type of sampling and procedure used to select sampling units
Probability, stratified and four-stage sampling; At each stage, a simple random sample of statistical units was selected. The statistical units are municipalities, cartographic blocks, households and adults. The selection of municipalities was operationalized with Python code (guarantees traceability of the random selection). The most important variable of statistical stratification is the geographical region, built with geographical criteria and the electoral behavior of its inhabitants in the presidential elections of May 29, 2022. The design of the sample considers the technical guidelines established in Law 2494 of 2025. The details of the formation of the geographical regions and the procedures used to select the statistical units are found in the technical document available for consultation by interested parties.
Sample size
Survey of 2,073 adults with the intention to vote in 2026, distributed in 54 municipalities in Colombia
Design margin of error and confidence level
The global confidence level is 95%. The overall margin of sampling error is 2.9%
Data collection technique
In-person household survey. Cards with photographs of the candidates were used as material to help the respondent. There was no compensation to the respondent for answering the survey. No data imputation processes were carried out
Names and surnames of the statistics professionals responsible for the survey
Samuel José Yaya Rodríguez.
Specific topics referred to
Voting intention for the candidates for the Presidency of Colombia
Verbatim text of the questions asked and the order in which they were asked
Consult the questionnaire used in the survey
Candidates, people or institutions for whom the investigation was conducted
Candidates for the presidency and vice presidency of Colombia. See the details in the questionnaire used in the survey
Geographic space and time period in which it was carried out
The survey was implemented in 54 randomly selected municipalities in Colombia. The field operation was carried out from June 8 to 11, 2026
Sample Municipalities
Agustín Codazzi, Barranquilla, Bello, Betania, Bogotá DC, Bucaramanga, Buenaventura, Cali, Canalete, Capitanejo, Cartagena de Indias, Cerro de San Antonio, Chigorodó, Circasia, Duitama, El Banco, Flandes, Florida, Fundación, Funza, Guasca, Ibagué, La Cumbre, La Virginia, Los Córdobas, Los Patios, Medellín, Montería, Neiva, Ocamonte, Ocaña, Pácora, Páez, Pasto, Pereira, Piojó, Pueblo Nuevo, Puerres, Puerto Boyacá, Puerto Tejada, Restrepo, Riofrío, Rionegro, San Francisco, San Jerónimo, San José de Cúcuta, Santa Sofía, Santander de Quilichao, Silvania, Soacha, Soledad, Toribío, Úmbita, Villavicencio
