Policy Perspective: Surveying | Column by Jorge Alcocer V.

Jorge Alcocer V.

In her closing speech, the official party candidate loudly boasted of her victory in the election campaign. Interesting: who did Claudia Sheinbaum beat? Like Xochitl Galvez, she was the only preliminary candidate.

I believe that the triumphalism of the ex-head of the capital’s government is due to the fact that in all published reviews she has a noticeable advantage over her main competitor. But this condition is not the result of his activities during the election period, it existed before, and it seems that there is even a tournament being held to see who will give him the greatest advantage. For example, at the end of last year Demotecnia gave it more than 50 points, and according to recent polls – from 20 to 30 points. I’ve only seen one “poll” that reduced that lead to less than 10 points. In quotes because this is a telephone exercise and does not follow the rules of probability sampling.

The use of polls as oracles predicting electoral victories is nothing new, much less exclusive to Mexico. Examples abound in all latitudes of the world. What I call “polling” is practiced among politicians and candidates. Worship the poll as long as it favors the one who calls it. Given this belief, I remember the warning of sociologist Maria de las Heras (qpd) who said:

The bad thing is not that the polls are manipulated in favor of a candidate, but that the person who pays for them ends up believing them.

Surveyors attribute many abilities to their goddess. Not only the actions of the oracle, but also the guidance of undecided voters until their participation and preferences are determined. The most recent case to support this view is the gubernatorial election in Mexico City last year.

The reasoning is this: almost all the polls gave the official party candidate a significant lead, which she proclaimed, like a broken record, “an irreplaceable advantage of 20 points.” In the end, he won by 8 points. Some commentators argue that the loud proclamation of 20 points led to almost half abstaining and that those who were undecided ended up voting for the clear winner. “With the champion until he loses.” This may explain the result in Mexico State; But other factors may have played a decisive role. To this day, I do not know of a single study on this issue.

The truth is that there is no conclusive evidence that polls have the effect that fans or mourners claim them to have.

For example, until March 2000, almost all polls gave the PRI presidential candidate a significant lead, which disappeared in the weeks leading up to Election Day until it was reversed. In 2006, the initial lead that nearly all polls gave to López Obrador disappeared as the campaign progressed. He knew this, but until the last day he declared that his own gave him an advantage of 10 points or more. The result was the closest in recent history.

Two days before Election Day 2021, Morena’s leader shared with me the results of the latest polls they commissioned for CDMX. With the exception of the mayor’s office of Benito Juarez, polls gave them a significant lead in the remaining 15 mayoral positions. As a result, out of 16, 7 won. The Morenists became offended that in every election campaign their candidates and leaders declare that their advantage in the polls is “irreversible.” These are surveyors who didn’t learn that even the lanes are two-way.

Surely there are those who believe that elections are won by opinion polls, or that today, as yesterday, repeating a lie makes it true. I argue that there is no evidence that polls win elections, and on the other hand, there is experience that can lead to the defeat of those who make the Delphic Oracle lose.

I am not surprised by the polls in favor of Claudia Sheinbaum. It would be surprising if the choice fell on Xochitl Galvez. To disqualify sociologists with extensive experience and recognized authority just because they do not support an opposition candidate is nonsense. There needs to be a calm and informed debate about the technical quality of polls, and about the transparency and ethics of election companies, especially their commercial relationships with parties, alliances and candidates. As in previous elections, at Voz y Voto we organize meetings between sociologists for this purpose.

INE also needs to improve its reporting of survey reports to impartially and objectively account for compliance with the law and regulatory standards applicable to survey companies. What amazes me is that, with few exceptions, companies report that they themselves fund the surveys they publish. INE accepts these statements as good, as well as the supposed “methodologies” some of them report.

INE does not evaluate or endorse survey results. But it will be useful to inform the public and public opinion about compliance with the law by election companies and the media in which they publish or disseminate their results. The fifth study report will be presented to the General Council this Thursday. I’ll get back to the topic.

Follow me on: www.vozyvoto.com.mx