If anything is a constant in Chilean presidential races, it’s the polls that make headlines in the run up to the elections. Disregarded by those who dislike the outcome, embraced by those polling high. Experts are seriously questioning the methods used by some of the leading polls and criticize their lack of transparency.
At the end of each week, famous polls like Cadem and Pulso Ciudadano launch the results of their weekly surveys into the media. Those media, national and international, don’t ask any questions about the surveys, so the people questioned and the methodologies used go unexplained, while the unvarnished results are “breaking news.” It’s a vicious cyle ahead of every election in Chile.
Just how credible are these polls? Why do they still get so much attention? And how much do they influence voters? Chile Today spoke with Marta Lagos (director of research organization MORI), Kenneth Bunker (founder of Tresquintos) and political scientist Claudio Fuentes to get some answers.
Cadem and its methodology
For Lagos, involved in Chile political polling ever since 1988, the main problem is the way surveys like Cadem operate. In their recent weekly polls, they use a group of 700 respondents. “You can’t call Cadem an adequate instrument to measure someone’s vote,” Lagos says. “International studies demonstrate you need at least 1,000 respondents to get credible results. And that’s the bare minimum. Also, their error margin is not according to standards. And we don’t know how Cadem asks its questions. If you ask someone, ‘Do you want world peace?’, all say yes. So, you can influence results by asking questions a certain way.”
In the latest survey released by Cadem, 24 percent of the Chilean electorate would vote José Antonio Kast, while 19 percent would vote Gabriel Boric. Fuentes, professor at the Diego Portales University, refuses to say this outcome is false, but he does point to the limitations of a survey like Cadem’s. “They don’t publish their database. It has limitations to correctly interpret the results. They often point at people from the middle and upper middle classes, living in urban areas, interested in politics, so that also influences the results,” he says.
Despite Cadem’s shortcomings, national media report every week on the results of each Cadem survey. The way they do that is misleading, Lagos says. “They interpret the results of Cadem as an anticipation, as something that is about to happen, instead of a snapshot of earlier intentions. Surveys never anticipate. The results we read come from surveys performed five days before, while people’s votes can change quite quickly,” Lagos says.
It’s not only media that are tied to the web of political surveys in Chile. In the case of Cadem, ties also run deep with the government. “There are few people in Chile who criticize the powerful model of Cadem,” Lagos says. “They surveyed for Piñera when he ran for president, they get paid by La Moneda. And now they get free publicity with their political polls, which is just a side business for these companies. Newspapers cherry-pick when it comes to surveys, they choose which survey they report on, without questioning background or methodology used.”
Need for regularization
Bunker’s website, Tresquintos, includes the results of all major political polls in Chile. He calls the surveys part of the electoral race, but emphasizes that he doesn’t trust them. “I think the biggest problem of these surveys is that it is impossible to measure the public opinion of people in a country that moves politically so erratically. There is no instrument that can measure everything that is happening in Chile. It goes beyond a social or political crisis: there are also things changing in Chile, like the political parties, the electoral system. And candidates also change, in their programs for example. For any survey that is impossible to capture,” Bunker says.
The whole sector of polling companies should be regularized, Bunker says. “I would suggest that all surveys in Chile have to comply with certain international standards to assure a better process. Right now, for example, we don’t know who finances the surveys.” Fuentes agrees. “The first thing has to do with transparency: surveys should start publishing their database, inform about their methodology used, everything should be public. The second thing has to do with the fact that in Chile, 15 days before the elections presidential surveys are prohibited. But still, private surveys are published, this must be better regularized. Maybe all surveys can get published until one week before the elections, but only those that comply with all standards.”
Influence on votes
An important question is how influential these polls are when it comes to the vote of citizens. Do people change their vote when they see their candidate is losing according to surveys? All experts agree this influence is minimal. “People choose a candidate based on what happens in their environment. They long lost confidence in the ruling class. Many don’t even vote,” Lagos says.
Fuentes points to the outcome of this year’s primaries as an example that polls don’t influence that much. “The electoral results of the primaries don’t coincide with the results from the polls. For example, everyone said Daniel Jadue would win the left-wing primary, but Gabriel Boric won. Same thing happened on the right-wing with Joaquín Lavín, who was favorite in all polls. In the end, Sebastián Sichel won.”
Editor-In-Chief Boris van der Spek is the founder of Chile Today.