The author makes the bold claim that opinion polls are of little use in predicting election results, and that in some cases, the majority of voters do not even make up their minds until the last minute. He cites the example of Senator McCain’s unexpected victory over Senator Obama in the 2008 presidential election, when polls had Obama in the lead

In this particular case, the polls were indeed inaccurate, which led the media to dub them the ‘polls are wrong’ election. As was pointed out in a subsequent editorial, the inaccuracy of the polls was due to the fact that McCain had the advantage of running against an incumbent president, whereas Obama had to run against an incumbent vice president. However, the author’s assessment that the polls are of little use in predicting elections is incorrect, as opinion polls do have some predictive power. For example, the polls in 2004 were within 3% of the actual outcome of the election, according to Nate Silver of The New York Times. Similarly, the 2012 presidential election was the closest presidential race in two decades and the polls accurately predicted the winner with 95% accuracy. This accuracy is likely due to the fact that, unlike in elections between sitting presidents, the partisan makeup of the electorate is consistent, and both candidates generally have the advantage of running against an unpopular incumbent. Therefore, the partisan composition of the electorate and the candidates’ popularity are more reliable indicators of election outcomes than opinion polling, which takes into account the fluctuating opinions of the electorate

Another fallacy in the author’s argument, is the assertion that opinion polls are of little use in predicting the results of elections. This ignores the massive amount of data that goes into these polls and the complex algorithms that calculate their outcomes. For example, in the 2012 presidential election, the Obama campaign invested over $1 billion on scientific polling, which took into account the opinions of over 15 million respondents. The 2012 election was the first election in which the majority of the electorate cast their ballots through internet polling, and the Obama campaign’s sophisticated algorithms were able to take into account how the respondents voted in different online platforms, such as Facebook, Twitter, and Google, as well as how they voted at specific polling stations. Therefore, the polling data used to formulate the predictions was anything but random

Indeed, the author makes several questionable assumptions in his analysis. He asserts that opinion polls only measure current voter preference, when in reality, polls are also interesting in that they reveal public opinion about candidates, political parties, and issues. Furthermore, opinion polls can be extremely useful in gauging voter sentiment about a candidate or party, and in gauging the public’s reaction to particular political events. For example, polls revealed that Republican presidential nominee Donald Trump’s racist comments about Mexicans and Muslims caused many Republicans to lose faith in him and pledge not to vote for Trump. This data was invaluable in forming a strategy to combat Trump’s appeal among Republican voters, such as criticizing his bigoted statements and appealing to his economic populism. Therefore, opinion polls are powerful tools that can provide us with valuable insight about voter sentiment and political trends

Finally, the author makes the assumption that opinion polls are accurate, whereas random guesses are worthless. This is a rather dubious assertion, as random guesses are notoriously difficult to make, even by amateurs. For example, many experts were unable to correctly predict the outcome of the 2008 presidential election, much less predict the winner in 2012. The 2016 election was no less difficult to predict, as polls differed widely in their predictions. Therefore, it is impossible to conclude that opinion polls are worthless, as the author suggests.