[ad_1]
How much have been trustworthy and I need polls of the 2020 presidential elections? You will receive a reply in a few days, when the general picture of the votes, those cast for mail, Those of early voting and those who attended on election day – will be final. For the moment, it can be said that once again polls have underestimated the electoral strength of Donald trump. As November 3 draws closer, most polls showed Joe biden widely favored and Donald trump with a lot of possibility limited. What happened in the hours after Election Day shows how this prediction turned out to be false.
Take the case of three states that have been under close observation for some time: Florida, North Carolina, Georgia. “FiveThirtyEight”, the site of Nate silver, which provides analysis and projections on electoral numbers, offered an average calculated on the most reliable polls in the hours prior to November 3. Biden was given in benefit, respectively, 2.5, 1.8 and 1.2. This turned out to be incorrect, especially in regards to the Florida, that Triumph won by 1.2 points in 2016 and which the president recovered this year, thanks in large part to the Hispanic vote of the county of Miami Dade – with an advantage of more than 3 points (to be precise, 3.4, but the count has not finished yet). This means that the margin of error, between the forecast and the reality of the vote, was at least 5.5 points.
A similar speech, even more sensational, can be done by Wisconsin me Michigan. The average of the surveys offered by “FiveThirtyEight” gave Biden ahead of 8.4 points in Wisconsin and 7.9 in Michigan. The reality was very different. In Wisconsin, at the time of writing, Biden obtained 0.6 percent more votes than Triumph. the Michigan it was awarded to the Democrat with a difference of 2.6. Another rather sensational case concerns theOhio. In 2016 the president had won it State quite easily (more than 8 percent) and this year its advantage was given by very small: about 1 percent. In fact, theOhio came back under control quite easily Triumph, about the same percentage as in 2016. There is no real hope of Biden and the Democrats then came from the vote of Iowa, or of Texas, as several polls predicted.
Alone Nate silver explained that once the counts are completed, the margin of error for forecasts Compared to the final result, it will be around 3%. Truth be told, this is not a great result. In 2016, errors were primarily attributed to the inability to truly capture voting attitudes of bianchi without a university degree, who had decreed the victory of Triumph in at least three essential states: Pennsylvania, Michigan e Wisconsin. Today, the reasons for errors may be different, depending on the states considered. For example, in the case of the county of Miami Dade in Florida, surveys did not reveal the important role that Hispanics (especially Cubans me Venezuelans) had in the victory of Triumph. According to some, this mainly depends on one thing. The greatest difficulty of To assign the electorate Hispanic than that White: more difficult to contact by phone, less likely to participate in Attention group, less present in the rallies that the president held in the state. However, in the end, that vote was in percentages much higher than expected. Triumph, deciding the fate of an essential state like the Florida.
Also in this, as some have pointed out, there is a tendency to classify sectors in the same box, for example “Hispanic voter”. population, cultures, conditions socioeconomic very different. The “Hispanic” voter no longer exists: they exist in the United States very different Hispanic worlds and this has not been sufficiently taken into account by various investigations. In the case of states of Midwest, the reason for the new bankruptcy seems more similar to that of 2016. I polls that is, they would still not really understand the weight of the supporters of Triumph. The reason has often been explained with the thesis of “timid Trump voters“That is, voters who they are ashamed of their membership and that, however, in the secrecy of the polls, they regularly vote for President. A pollster fromTufts university, Brian schaffnerInstead, he made a different hypothesis. And it is that the voter of Triumph has, like its president, a natural distrust of media and world academic, and therefore less likely to answer a survey honestly.
The problem, some point out, is that in general it seems difficult to map and represent phenomena populists. The polls failed in the case of the vote on Brexi, in 2016; and yet they failed to predict the emergence of right-wing populist candidates in Australia. However, the repeated failures of many forecasting models, today as in 2016, have given a new voice to those who ask to reconsider the entire system. He wrote Margaret sullivan in the Washington Post: “We will no longer have to place the public’s full trust in polls and those who interpret them, as we have become used to. The survey today appears as something false. Or at least it’s our way of taking it so seriously. “
Eventually, the annoyance Because the new fiasco and the request for a profound renewal also come from the world of research institutes. In an interview Chris Kofinis, CEO of Park Street Strategies, a Democratic pollster, spoke of a “dramatic collapse” of polls and pollsters, “who are simply not doing their job”. Caffeine Talk about focus group wrong, with the researcher trying to impose his opinion on the people interviewed; and it also mentions the economic reasons behind certain analyzes. States like Texas, Ohio, even the FloridaThey’ve never been seriously at stake, but make it look like “battlefield“It creates interest, requests new research and, therefore, causes an economic investment on the part of the parties and the media from which the research institutes benefit. The problem, however, is that the system now seems to have come to an end. If the inability to predict the increase in Triumph in 2016 it was explained by the difficulty of interpreting a new phenomenon such as Triumph, the replica in 2020 has less justifications and somehow it is more serious.
“The crisis of polls is a catastrophe for American democracy ”, he titled“ The Atlantic ”. And for “Slate”, “the problem is not that the polls are wrong, the problem is that they are unnecessary“The question is whether and how much it will really change. bill still ongoing, it may be too early to understand what really went wrong and what can be done to avoid another failure. Reflection must surely also involve the world information, which has often abdicated an essential function, that of excavating the guidelines electoral, now entrusted to numbers me polls. Wrote, again in the “Washington Post,” a political analyst like Henry Olsen: “Political prediction is an art like a science. Pollsters and analysts, myself included, have clearly counted too much on the scientific side this time. Next time, we need to rediscover the piece artistic, e electoral instinct”.
[ad_2]
