Polls

Polls were a leading figure in the past presidential election. However, instead of being an instrument of measurement and a source of trust in the results, they acquired an unusual and highly destructive transcendence. They ended up being means that manipulated the election and impeded conferring certainty on it. In the light of this, it is evident that calls will be issued to regulate the service. I beg to differ: what are missing are opening, transparency and competition.
Is it possible to manipulate an election with polls? According to the Dictionary of the Royal Spanish Academy, to manipulate means “to intervene with clever and, at times underhanded, means in politics, in the marketplace, in information with distortion of the truth or justice, and at the service of special interests”. The key element of this definition is the conscious seeking of a determined objective: to distort. The evident question is whether it is possible to utilize polls to manipulate the voter. According to Edward L. Bernays, an expert in propaganda, “the conscious and intelligent manipulation of the habits and opinions of the masses is an important element in democratic society”. In other words, this is an exceedingly subtle issue in which the crucial factor lies in the objective: if this is about intentionally distorting the information, we find ourselves before a case of manipulation; however, if it concerns attempting to exert an influence on habits and opinions, it does not. The main point is not, therefore, whether the polls influence the voters’ behavior, but rather whether they were utilized consciously and deliberately by those who produced or published the polls with the express objective of altering or inducing perceptions among the citizens.
There are two key issues here: the difference between the last polls published before the election (there is a five day ban prior to election day) and the final result; and the manipulation that the polls may have been used for. These are distinct themes that merit a differentiated analysis because they are not the same.
One great myth of the election resides in the assumption that the margin of advantage of the leading candidate with which the process initiated would be maintained until the end. Anyone who has observed elections around the world knows that elections always narrow, generally between two contenders. The exceptions to this rule have perfectly obvious explanations. One exception was the election, in the second round, between Jacques Chirac and Jean-Marie Le Pen in France in 2002. Instead of the traditional Left-Right confrontation, the election ended up being between the Right and the Extreme Right, which led to the Left voting for the Right. Chirac swept through with more than 80% of the vote. Another similar case was that of Eruviel Ávila in the State of Mexico a year ago: there the PAN as well as the PRD were over-confident, supposing that the PRI would nominate someone within the close circle of the then Governor Peña-Nieto and thus they did not construct a candidacy of their own. In fact, they essentially bet that Eruviel himself could be the candidate of an opposition alliance. The result was not as overwhelming as that of Chirac, but nearly.
All elections close in the end and many of the polls do not achieve capturing the evolution that these register in the voter in the last days prior to the election. This year’s election ended up defining itself by the anti-PRI vs. the anti-López-Obrador vote and this led to the PAN’s falling much below what the polls had said: this is a typical pattern and has nothing to do with any manipulation. The voters adjust their preferences according to their wishes but also according to their fears, factors that clearly modified the final results with respect to the last polls published before they were legally banned.
Much more complicated than this is the issue of the alleged manipulation. In Mexico there are many types of polls, many financed by parties and candidates, which are published as if they were a science, when on many occasions they reflect sampling frameworks constructed ex profeso to promote their clients. Others simply reflect the lesser quality and technical skill of the pollsters. The way that the pollsters treat the undecided and non-responders has great influence on the way the results are presented. The point is that the diversity of available polls makes it easy to see them as similar and comparable, when in reality many clearly are not. That is, some constitute expressly designed attempts to manipulate the voter.
The great novelty of this presidential race was the appearance of a daily poll, something that had never been present before in Mexico and that, as far as I know, does not exist elsewhere in the world. A daily poll that shows as many as eight-percentage-point shifts in just one day is immediately suspicious. The poll can be technically impeccable, but its appearance every single day lends itself to all kinds of interpretations. Worse still when the way of presenting the daily news in the same newspaper and, on some cases, the opinions of their columnists, coincides with the poll’s results, all of which generates an inevitable sensation that there is something fishy behind it and undivulged connections between one and the other.
The question is what to do about this. It is impossible to determine whether there was an express attempt to employ the polls as a means of manipulation. However, the issue is quite serious in a country that has yet to achieve the basic elements of the democratic process, i.e., recognition of the election result, which is why the theme will inexorably return to the political discussion.
In the face of this, I anticipate that the typical summons will be issued to prohibit or regulate polls, sanction the media that publishes these or, in whatever case, continue deepening the prohibitionist regime that is so harmful for the development of a democratic country. And worse still because nothing would be achieved, as has been demonstrated in the case of the campaign finances.
Another answer, very distinct, would consist of obliging the pollsters to publish their historical record, that is, to make public in each poll the comparison between their predictions in previous elections and the final result. The electoral authority could emulate the way in which the consumer is warned by the health authorities about cancer in cigarette boxes with a clear and unmistakable label with a caption such as “this poll was off by X number of points in the last election”. With these data, the voter would have the necessary information to be able to discriminate between the polling professionals and crass connivers.
There is no perfect solution but the degree of conflict that Mexico is experiencing compels us to think of creative means that will lend certainty to the process. Competition is the only possible solution and this entails opening and transparency. This would also help to advance toward electoral processes that are completely legitimate, exposing the charlatans and professional string pullers.