Election polls are inherently open to opinion, especially on the one hand kitchen which may end up taking on a new meaning (now, at the IEC, it already has seven) as a process of elaboration with the salt and pepper of demoscopy, and by a hermeneutics that must be the most important after the biblical one. Those responsible are the donkey of the critics' blows, for example, the tireless director of the CIS (Center for Sociological Research), José Félix Tezanos, a professor at the University of Santiago with numerous publications, who is criticized survey after survey. But for the moment, it's holding up; Jordi Muñoz, also with a spectacular academic resume, was not so lucky; he only lasted three years at the CEO (Center for Opinion Studies).
Subscriber Aniol Fusté raises the issue of the polls with me at the ARA headquarters, emphasizing that "when they tell us about the results in Catalonia, they list all the 'possible' deputies (whether in Parliament or Congress). That is, they don't usually separate them by constituency."
"I understand," he continues, "that it's the same newspaper that commissions these surveys from companies that carry out the studies and that, surely, a detailed result separating Barcelona, Girona, Lleida and Tarragona is more expensive. Wouldn't it be, but, fairer? We live in a country where absolute majorities have been neither seen nor expected for years and, therefore, the results of the medium and small parties win. If we put everything in the same bag, we lose the perspective of knowing that the seat that is dancing in Lleida could be taken by this party or that other. Political parties do have detailed surveys by constituencies, provinces, regions and municipalities (and so they decide where they need more or fewer events). If they can do it, it shouldn't be so difficult."
He concludes that separating by district "would contribute to a greater understanding of the lists in Girona, Lleida, and Tarragona and would help the candidates in these three areas earn votes with greater focus and attention (and therefore, with greater journalistic control and oversight). If we say that the entire territory is important, we have to practice it."
Deputy Director David Miró, in charge of politics, responds to subscriber Fusté. "The issue of polls and constituencies," he states, "has a technical component that I don't know if I can explain, but as a journalist, I fully understand your request. In fact, when the company we commissioned a survey from presents the results, we always ask: can we provide the results by constituency? We do so because it has undeniable informative value. To offer the results by constituency with a reasonable margin of error, we would have to conduct four different surveys in the four territories, with 800 responses in each, and that is beyond any budget.
The only thing that can be done, Miró concludes, is to fine-tune the models to extrapolate the results for Lleida, even if you have few surveys. We use a hundred of the 800 needed to conduct a reliable survey across all of Catalonia. But pollsters are reluctant to offer results with high margins of error because it undermines their credibility. In any case, the reader can rest assured that we always insist and trust that in the future it will be cheaper and easier to offer these results."
Given such clear evidence as the advisability of reporting results by constituency, the optimism of the will leads this Ombudsman to believe that this should eventually be possible, whether by optimizing supplier services or through sponsorship.
Algorithms and the Spanish flag
Reader Pol Rodoreda sees an ad for hearing aids in the digital edition with a Spanish flag in the background. Naturally, he complains, because flags are symbols that are both powerful and very personal: "I think it would be interesting, as a Catalan newspaper, to have more exhaustive control over the advertising that appears in the digital edition."
Deputy Director Ignasi Aragay explains: "There are ads that are not directly contracted by the newspaper. Like the case reported by the reader, these are programmatic digital advertising, an automated purchase of space that connects a brand with a supposedly similar user. For whatever reason, the algorithm has established an inappropriate link. When we receive justified complaints, as a newspaper we ask that these ads be removed for our readers, as we have done in this case."
Digital advertising is experiencing tremendous growth, even on platforms that overlay movies and series with ads and even include static advertising during breaks. Innovation in any field implies a lack of control and legislative loopholes, and in this case, we'll investigate whether the invasion of algorithms is likely, in some cases, to violate Article 18 of the Constitution, which guarantees the right to privacy and confidentiality of communications and limits the use of information technology in this delicate environment. Privacy policies and cookie regulations are a first step, but in such a fluid ecosystem and with hackers with highly developed computer skills, circumventing the law is easier than in other areas of law.
Finally, algorithms don't capture intentions, so—as an example—under the cover of a recent investigation into Franco's regime, the dictator kept appearing to me fifty years after his death. It wouldn't be surprising, for example, if you're interested in Catalonia and its circumstances, that reverse algorithms wave the Spanish flag.
The Readers' Ombudsman pays attention to doubts, suggestions, criticisms and complaints about the contents of the newspaper in its digital and paper editions, and ensures that the treatment of information is in accordance with the codes of ethics.
By contact the Readers' Ombudsman You can send an email to eldefensor@ara.cat or record a message of no more than one minute on WhatsApp at 653784787. In all cases, identification with your name, surname, and ID number is required.
