Opinion polls only good for entertainment
Surveys are seldom to be taken seriously, Kaleva observes with the US election in mind:
“The failure of the polls in the US was explained with the excuse that many people who voted for Trump didn't want to say who they would vote for in the polls. Coming out as a Trump supporter was perhaps less socially acceptable than saying you were a Clinton fan. … In 2011 the populist Finns Party won an equally unexpected landslide victory in Finland. … The failures [of the polling institutes] have prompted them to introduce diverse correction factors to improve the validity of their predictions. Sometimes these factors increase the accuracy of a forecast, but sometimes they lead to even worse results. … At best surveys provide an indication of the people's mood that is useful for reaching a decision. But in many cases their entertainment value is higher than their value as a reflection of the reality.”
Journalists refused to face the truth
The fact that many journalists were taken aback by the results of the US elections has to do with the basic bias of the media, Le Temps believes:
“In a country where each camp receives half of the vote, the vast majority of media represent only one side. ... How to claim that you're indispensable for democracy, as the media rightly do, when you represent only half of public opinion? Two phenomena result from this disequilibrium. The first is the public rejection of traditional information sources that are invariably losing their customer base to the Internet, which is considerably less reliable but more diversified. The other is the rise of populism. Because what it considers the 'freed word' is a response to the hijacking of public discourse by the establishment. The latter, so the idea runs, has little interest in the concerns of large population groups, many of whom already feel economically or socially disadvantaged.”