Polls still failed | The surveys present …

Telephone polls and political operations caused a real catastrophe this Sunday. Those who made a big difference in favor of Fernández – Fernández were four consultants who predicted advantages of nine, ten and even twelve points. They focused on conducting household surveys, that is, interviewing respondents at home. This is an expensive method, but one that allows for a better approach with the interviewee and, most importantly, reaches the modest sectors and young people. The good addresses of PASO come from pollsters based on telephone surveys and constituting the vast majority. Good part, worked for Casa Rosada. The method, known as RVI, is automatic and provides quick access to thousands of people. It's cheaper, but today, there are few landlines, they live in middle-clbad homes, young people do not respond and those who respond, in general, are older people . "The face-to-face and the phone are two different worlds and two different outcomes," said one of the consultants diagnosed during the last week of the election campaign.

The three pollsters who worked for the Fernández – Fernández formula were Roberto Bacman, Hugo Haime and Analía Del Franco. They formed a team between the three and the base of the Bacman – Haime – Del Franco polls. On the orders of the campaign command, they were forbidden to broadcast the results of their studies. The journalist had access to the data and recorded differences, even though the condition was to maintain confidentiality.

The other consultant who showed a large advantage of FF over MP is the Spanish Alfredo Serrano, from the Latin American Center for Geopolitical Strategy (Celag). He broadcast an investigation that gave Fernández – Fernández nine points distance Macri – Pichetto. The distance was much greater – almost 17 points – but it broke with the weather conditions installed by Casa Rosada which spread, to the cover of Clarin, an investigation that predicts a virtual tie.

There have been other consultants who have been using the system called IVR, Interactive Voice Response, a system that produces thousands of calls per minute and in which questions are answered by typing numbers on the keyboard . The IVR system is focused on fixed lines, so it over-represents the middle clbad and the over-60s, who are usually the only ones to answer the phone. To top it off, the system is somewhat saturated, because there is a lot of publicity, or even a lot of Together, for the change, the level of rejection is high, especially among young people. The pollsters who manage this system weight, which means that the ability of the professional to replace the lack of small and young sectors is very heavy. CNN has just published an internal regulation – although it has been made public – that it will no longer publish surveys of the IVR system.

The truth is that most of those who used the phone method recorded a virtual link situation or, at most, four points of difference in favor of FF. The stakes were not small: the government launched a mbadive offensive on the basis of these polls and was crowned on Friday when it was announced that a consultant, Elypsis, predicted very little difference for FF and elections proportionally positive for the party in power. There were consultants who used the IVR and who said Friday that there were four points of difference in the province of Buenos Aires and that Macri-Pichetto had triumphed in Santa Fe, in the province where the Todos Front had won by ten points. A good part of the pollsters who argued that there was a hidden vote for Peronism, but one of the challenges of a consultant is to "fish" what is wrong. it happens to those who are elusive to answer or insist on being undecided.

Bacman – Haime – Del Franco, through face-to-face surveys, made a difference of 12 points in the province of Buenos Aires. This gave a base for the national scene. The face-to-face method also has obstacles. For example, in the buildings, it happens that the porters do not let in and in private quarters, there is no access. This part has been supplemented by another method, called Computer Assisted Telephone Interviewing (CATI), in which the system randomly calls fixed lines, but it is not a machine that performs the survey, but an operator. This allows for more and better data.

One of the most serious problems that existed with face-to-face surveys was that the pollster had to write, then that form was sent to the system and everything was heavy. It is now supplemented by tablets or mobile phones that monitor the sounder and, in addition, immediately transmit responses to the central system.

Methodological surveys include cell phone surveys. The problem is that there are only user bases that have contracts, not maps, which greatly limits the arrival in the modest sectors. There is also the immense difficulty that the person feels invaded when she calls her mobile phone and the level of rejection is very high. Finally, there is the disadvantage that the person who has a CABA mobile phone can be in Entre Ríos, ie it is very difficult to locate it socially and geographically. This is not the same as the visit of the interviewer, who sees the house and does the interview there.

The truth is that Bacman – Haime – Del Franco, with mainly face-to-face investigations, had a considerable advantage for FF, but Nicolás Trotta, rector of the Metropolitan University of Education and Labor (UMET), coordinator in the field of investigations of the campaign Fernández – Fernández, he agreed with the consultants that the results would not be divulged to avoid the triumphalism. In addition, it was risky and even difficult to present obvious benefits in relation to the offensive triggered by the investigations broadcast by the executive in the mainstream media.

The trio of pollsters of the FF also conducted a survey in the conurbano with ballot boxes and ballots to determine if there really was a reduction in favor of Maria Eugenia Vidal. Respondents received the ballots and a ballot box was made available to them. The conclusion was that the cut was very minor. Kicillof has not lost votes regarding Alberto Fernández and Vidal had three voting points for Lavagna and Espert. Nothing that moved the result. This was confirmed Sunday: Vidal took 32% and Macri a little over 29 years old.

Beyond surveys, some data flowed from logic. CFK arrived with a 35% voting intention, Alberto joined then Mbada, which gave a reasonableness to the 42 points that emerged from the results of the poll. On the other hand, the positive badessment of Macri not reaching 35%, so it was not reasonable to increase even more when voting. As Serrano pointed out, "the dominant slogan of this election was that Macri does not follow, that the macroeconomy must be ended".

In the past few days, what has been called follow-up has come to fruition, that is, investigations have continued to determine whether there is a change in trend. The benefits were always around 10 points, sometimes a little more, but the policy remained the same: do not spread or encourage triumphalism.

Throughout the day Sunday, a lot of work was done – the pollsters near the schools – endorsing the findings. Finally, a study of 2,000 witness tables yielded the figure of 14 points of difference. Nevertheless, a relatively low profile was maintained, filtering that the distance was seven or eight points.

Of course, the general conclusion is that polls failed on Sunday. It turns out that the media scenario was dominated by those broadcast by Casa Rosada and those made with the IVR. The result produced a cimbronazo, because there were not four points but 17. Of course, the document will raise strong controversies on the methodologies.

[email protected]