Mexico.- In an election you cannot get a vote in the ballot box without understanding the voter who casts it. Election polls can help you find out what potential voters think or feel. A well-designed survey can help gather information on customer satisfaction, they can also help know the intentions and opinions of the voters, which means that a candidate will have a better understanding of how to craft their policies and platforms to meet the needs of citizens.
This allows them to evaluate both the political landscape and the viability of a candidacy, with which the results of the polls become the photograph of the political moment that a city, state or country is experiencing.
However, the results are not always presented accurately to the public, some media give preference to candidates who in reality do not have the opportunity, showing figures of vote intention very far from the real result, and even, when the results They do not favor the person who hired the services of the pollster, it is decided not to publish them to the public opinion, or else, there may be obstacles for the pollster when their results do not favor a candidate.
In an interview for Debate, Heidi Osuna Peraza, director of the Enkoll pollster, spoke about what really gives citizens certainty and credibility when talking about polls, since a pollster “must be accurate and responsible when publishing numbers.”
With 15 years of experience in opinion studies, six of them in charge of her pollster Enkoll, Heidi Osuna Peraza has positioned this pollster for her frequent successes in the final results in electoral processes.
In 2018, for example, Enkoll was one of the three most successful polling houses with respect to the presidential election and, in the same way, in the results of the head of the Government of Mexico City.
In 2021, they also obtained several hits, one of them was in the results of the elections in Sinaloa, when a week before the electoral process they published that Morena’s candidate, Rubén Rocha Moya, had 54 percent preference; While the PRI candidate, Mario Zamora, had a 31 percent preference, the official results of the INE being 57 percent for Rocha Moya and 32 percent for Zamora.
“Instead, some national media made an appearance of a closeness of the entire election, that closeness never existed, there was always a wide difference between Morena and the Alliance in Sinaloa,” he commented.
In the case of Sonora, they published that Morena’s candidate, Alfonso Durazo, was going to win with 51 percent preference and in the end the INE confirmed his victory with 52 percent; while his closest opponent, the candidate of the PAN-PRI-PRD alliance, Ernesto Gándara, had a 33 percent preference and in the end obtained 35 percent. Results within their margin of error, and the same thing happened, some media, in some polls, had a technical tie or a minor difference between Durazo and Gándara, when the difference was actually quite noticeable, explained Osuna Peraza.
In Nuevo León, Enkoll gave Samuel García, candidate of the Citizen Movement, as the winner, with 39 percent and a 37 percent preference; and his opponent, Adrián de la Garza, candidate of the PRI and PRD alliance, had it with 26 percent and was left with a 28 percent preference in the official results.
“Here several pollsters brought the other way around, they brought the PRI in first place and Samuel García in second, and some other polls that we do not publish because our client was not the one who was winning, so they were internal polls for some political parties; but, well, in 2021 we repositioned ourselves as an accurate and responsible pollster when it comes to publishing numbers ”.
Find the informant
In the words of Heidi Osuna, the secret to obtaining these results, very close to what the public really thinks and thinks, is to ask in the places where the housing surveys work, since some other factors must always be taken into account, such as the rejection rate, since in some households they do not open the door or do not give the opportunity to respond, so it is vital to try to find that informant, try to go to the houses at the times when it is most likely to find the informant .
“Otherwise, you only reach students, retired people and sometimes you tend to exclude people who are in the workforce, as they are usually away from home all day. Also have very specific quotas for socioeconomic levels, there are socioeconomic levels that sometimes it is very difficult to reach via the field, so if you do not put these quotas you can have an overestimation of certain political parties, such as Morena, or you also have to measure, perceive if the population is afraid to answer the survey, if they are not giving you information, because that also usually happens ”, indicated the director of Enkoll.
Another strategist is to record the interviews and have the pollster geolocated because the polls are done through a mobile device (tablet) and thus from the office it is possible to give specific follow-up to each of the pollsters to know that they are in the section that they were indicated as a sample and also that they are applying the questionnaire correctly.
“That there is no vice on the part of the pollster when it comes to approaching the interviewee,” he explained. And finally, and the most important, is to be clear, to note that Enkoll does not change numbers. “If we have a client and the number is not favorable, we do not publish the survey, but we only promise not to publish that number; but, apart from that, we have to publish what people are telling us, “he mentioned.
When they don’t open the door
With the arrival of the covid-19 pandemic and health restrictions and safe distance recommendations, pollsters also had to adapt for some time and make use of the telephone survey option.
In this sense, Osuna Peraza explained that also in this case there are places where this type of survey works better because it is possible to reach those high economic levels, which sometimes with the housing survey you do not have access.
“If you want to do a survey in a place like San Pedro, Nuevo León, it is better to do it by phone, because it is very difficult for more than 40 percent of the population that exists in that municipality to open the door in that municipality, and there are other municipalities or other states where there are not such high peaks in the economic levels of the top of the pyramid ”, he said.
Also, at times, when conducting telephone surveys, higher levels or more educated people may be overestimated, but there are states, such as Sinaloa, in which a telephone survey can be useful, but, above all, if it is made with an interviewer and not via a robot, because, for example, that robot cannot verify that the age is real, that a woman or a man actually answers, and also if that person works or lives at that address, he explained.
“We make a mix between housing and cell numbers and the same thing, a robot cannot check if a child is answering it or not,” he added.
Another important fact is that, in order to overcome the mistrust or little credibility that the interviewees may sometimes have, more by telephone, it is that the surveys are as short as possible, as this reduces the percentage of rejection or abandonment of the survey, he noted.
One fact that stands out is that for the pollsters in Mexico to consider the interviewee as a person of high or medium-high socioeconomic level, it must be a house with at least two bedrooms, two bathrooms, there must be at least two vehicles in the house and that the informant, the head of the family, has completed university studies. Although in Mexico, despite having the above, only 1 percent of the population is considered to be of high socioeconomic level, said Heidi Osuna.
The photo of the moment
Knowing the socioeconomic level of the informant is very useful, since certain proposals that the candidates have are focused on a certain group of the population, although he clarified that the greatest impact is on the gender and age range of the interviewee.
“Only at the extremes is where you have these great differentiators in the lower class and in the upper class that, if you are campaigning to give social support, then to the upper class it is something that is not a great incentive, but for the lower class it’s likely to be a great incentive, so it does work, ”he explained.
In this case, Osuna Peraza mentioned that the National Action Party is almost always located from the middle class upwards, Morena is practically at all levels, although with few in the upper class, and the PRI is also present at all levels. , but already to a very small extent.
Heidi Osuna asserted that citizen participation in this type of survey is of utmost importance because, through surveys, a photograph of what is happening at that time in their locality, city or state is obtained. He recommended not answering interviews that are aimed at giving the result that they want, that is, participating and answering those surveys in which their opinion is really asked and not in those that are influencing their response.
“The intonation of the pollster has to be exactly the same for all political parties when they are questioning a candidate,” he exemplified.
In addition, the pollsters who want to favor candidates can be identified because they regularly call at inappropriate times, present themselves as part of the candidate’s team, and in the questions they usually look for an answer that favors them, such as: “’This candidate has done such bad things, would you vote for him? ‘ Then you can see that they are wanting to lead you to a specific result, “he said. So, if in the survey they begin to speak ill of a political party or a candidate and the question is not seen as neutral, that is where you have to be alert, he recommended.
Read more: Who accompanied AMLO in the Cry of Independence?
For the interviewers in Mexico to consider the interviewee as a person of high or medium-high socioeconomic level, it must be a house with at least two bedrooms, two bathrooms, there must be at least two vehicles in the house and that the informant, the head of family, have completed university studies.