A few days ago, the third survey of Morena militants and supporters to renew the national presidency of that party concluded.
The polls were necessary since the Morenista groups could never agree on the method of choice. In the exercise, Mario Delgado, former coordinator of Morena’s bench in the Chamber of Deputies, was victorious, with supposedly 58.6% of the preferences against 41.4% obtained by Porfirio Muñoz Ledo, the veteran politician who in the past was president of the PRI and PRD. The “confidence interval” of the survey is an implausible 2.3%. There is the heart of the matter.
Examining the report produced by the pollsters Parametría, Covarrubias y Asociados and Demotecnia 2.0, the title of the book How to lie with Statistics comes to mind.
It is a case of electoral alchemy so blushing that it would not be surprising that the courts throw back the entire exercise in order to return to the starting point.
Let us remember that the survey was necessary because Morena’s membership list was never clear. When sympathizers were recruited to form the movement-party care was never taken to store the data and keep it up to date. Once a Morena president and Yeidckol Polevnvsky were elected as general secretary, but from then on the party functioned under the benevolent guidance of the benefactor in the National Palace.
Assemblies were held in the different regions, but all important decisions were made by him (that is why the registry was totally superfluous).
Candidates for popularly elected posts were raffled to prevent “ambitious” candidates from emerging in Morena. It seems a very healthy principle that curiously was not applied to also choose the candidate for the Presidency of the Republic.
When Polevnsky Gurwitz advanced to interim president, something that should be temporary, they realized that: one, they were no longer going to get rid of her, and two, that without the registry it would be very difficult to convene a National Congress to renew the leadership.
Polevnsky Gurwitz, for his part, never rushed the preparation or took care of purifying the list of militants, which was contested before the courts by those who discovered that neither all who are nor are all who are, which in a A party that calls itself a bulwark of democracy would have been elementary.
In other words, convening a congress-council-assembly to reach agreements became a tortuous process of demands and counterclaims that finally culminated in the VI Extraordinary National Congress, which agreed to appoint Alfonso Ramírez Cuéllar as “interim national leader.” .
A few weeks later, Ramírez Cuéllar sued Polevnsky Gurwitz for a fraud of 395 million pesos, a process that is still ongoing.
In the end, the court ordered Morena, in the absence of a registry, to carry out a survey to renew the leadership of the party. Thus we come to the first, second and third surveys.
The electoral alchemy of the third poll begins there: where the data is not reported in the press in a transparent and easy to understand way.
In reality there were three candidates: Delgado, Muñoz Ledo and the abstention. It turns out that 34.1% of those surveyed did not vote for either one. A little more and abstention wins.
Delgado actually obtained 38.6% of the preferences and Muñoz Ledo, 27.3%. The difference is reduced from 17.2% to only 11.6%.
There is more: the pollsters asked the participants if they knew that Morena was in the election process, but 56.6% did not know. Still they continued to participate in the exercise.
If we have almost 60% of respondents who do not even know what it is about and among all of them there are 34.1% who do not choose either of the two candidates, where does the magic confidence interval of 2.3% come from? The survey is deficient from its very origin. The companies should then sample Morena’s militancy, preferably, including sympathizers. If in Mexico City, for example, there were 50% of Morena’s membership and 50% in the rest of the country, a sample would have to be selected respecting that distribution of members.
But what emerges from the report of the three pollsters is that it was an apparently random sample according to the number of voters in certain “sampling units” (districts) that were only categorized as urban, rural and mixed.
In an extreme case, very little could be sampled in regions where Morena has many militants and much in regions where they are scarcer. The survey does not represent at all the collective will, so to speak, of the militants.
In the second poll, concluded two weeks before the third, Delgado and Muñoz Ledo tied for percentage. That in two weeks a gap of 17.2% has been opened between the two (actually 11.6%) is not feasible in such a short time.
The only thing that shows is the extraordinary margin of error that both surveys had, which in no way can be 2.3% as the polling companies would have you believe.
It is enough to consider the following: if almost 60% of those surveyed did not even know that a new leadership was being elected, the most likely thing is that they were guided by the recognition of the person’s name. In the first survey, Porfirio Muñoz Ledo obtained 41.7% of name recognition and Delgado only 27.1%, and a few weeks later, although 60% do not know what it is, a majority select Delgado?
Not even with a thousand spectacular can this effect be achieved throughout the country. The only thing that the numbers show is that the variance of these surveys is enormous, that is, the confidence interval is such that the votes of Delgado, Muñoz Ledo and even abstention overlap.
Dear pollsters: it is obvious that your third survey has a very high margin of error, because: one, it does not respect or consider the distribution of Morena militants in the country; two, because the majority of those surveyed do not know what it is about, and three, because a third abstained.
The margin of error in this exercise is enormous and I would not be surprised if the roles were reversed in a fourth survey. That the confidence intervals of the percentages obtained by Delgado and Muñoz Ledo do not overlap is, simply and simply, a hoax.
If the survey reported had been conducted by an intern in the field of statistics, he would have failed.
And a disclaimer: I know both candidates personally and I do not doubt their good will. In this case, it would be better if they sat down to talk to reach an agreement. It would be much more convenient than relying on surveys that are not representative and carry a very high margin of error.
The fault lies not with the INE, an institution that has been very important to Mexico, but with Morena, who has not been able to order its registry or define clear rules of internal operation. The one who is pleased with so much confusion is Polevnsky Gurwitz.