On July 2, the president of GAD3, Narcisco Michavila, savored the success of municipal and regional governments on Twitter. He affirmed that he had only deviated in Madrid and Barcelona and completely failed in Ourense. Michavila, who described the generals as "much easier", championed the dominant trend that PP and Vox would add an absolute majority. He clearly failed. The demographic bubble of recent weeks has burst. Instead he was right the thesis that he maintained almost alone another company, 40dB d, which maintained that the right would not obtain the absolute majority. He also skidded the Center for Sociological Research with his majority for PSOE and Sumar, although in the last survey he was one of the least erroneous.
The small sample, frequently of 1,000 interviews or not much more, with which polls are carried out for all of Spain in elections that are actually 52 independents, one pro-province, is an essential deficiency. In the campaign, when tracking proliferates, daily surveys, in the main companies the samples grow considerably. Even so, the 8,001 interviews of the GAD3 on Monday of last week represent 154 interviews per demarcation.
Nor was there in the huge wave of surveys that led the popular Núñez Feijóo to Moncloa at the hands of the ultra Santiago Abascal an adequate assessment of the Spanish electoral trajectory, important for the kitchen, the calculation of the behavior of those who do not know or do not answer.
With a certain demoscopic Adamism, there was amnesia about the resounding error of attributing a majority to PP and Ciudadanos in 2015. It was not taken into account that at most one bloc had fallen seven short of the barrier of 176 seats, the one on the right in 2016, with 46% of the votes, which set the bar for an absolute majority perhaps at 47%. This is a percentage that the bloc of the right has never exceeded in Spain, except in 2011 if UPyD is included in it. There are signs of a shift to the right in society, but not with the intensity necessary to break the historic majority of the combination of center-left and nationalism.
Before, the PP could have an absolute majority even with 42%, but now much more is needed. For the prediction to be fulfilled, a very low participation was required that was not announced. In the end, the popular and Vox added 45.4% with which the 169 deputies of 2016 repeated, with seven tenths less. It is thus seen that there is no exact threshold of the dear absolute majority.
Another fundamental element consists of what can be called the demoscopic consensus, the confluence, especially towards the end, of the majority of the predictions. There, Narciso Michavila acted in the vanguard. Yesterday he attributed the errors to not properly assessing the fear of Vox.
Of course, the absurd ban on publishing polls from six days before has its weight. In this campaign in the last week there was an involuntary, but determined effort by Feijóo to mobilize the progressives, with a more than improvable performance.
However, on Sunday in its survey for Mediaset, GAD3 returned to give the absolute majority to the right, with 181 seats. Sigma2 gave them 173 and Metroscopia, 177. The 40dB ones from the last week, which were not published, continued in the opposite thesis. This company, directed by Belén Barreiro and who works for El País and laSer, had the best performance by keeping PP and Vox below 176 seats. And if the percentage of votes in the polls on the last legal day of publication, Monday the 17th, is analyzed, it is the one with the best record, with a deviation of 5.6 points. The CIS flash survey had 5.9, compared to 8.5 for GAD3 and 8.6 for Sigma2. That of Ipsos for La Vanguardia, on July 9, deviated by 6.8.
As Enric Juliana maintains, in Spain there is currently a demoscopic bubble that burst with a roar.
