2020 PRESIDENTIAL ELECTION – ARE THE POLLS RELIABLE?

As I write this, there are 102 days until Election Day, November 3.  102 days until we will be voting in what I and many other people consider the most significant election in our lifetimes.   Unlike most elections, this year voters will have a clear choice.

I believe voters will be deciding the future of America for some time to come.  On one side is capitalism, free enterprise, law and order, self-determination and a personally unpopular president whom about half the country dislikes intensely but who has demonstrated that he gets things done.   One the other side is socialism, crime, chaos, a government that heavily controls how we live, and a “nice guy” of questionable mental capacity

In recent months the presidential election polls have shown Biden, the latter choice, to be leading President Trump, the former choice, nationally by between eight and 14 points depending on the poll.  Those same polls also have reported that Biden is leading in most of the battleground states, in some cases by double digits.  Taken at face value these polls appear to be predicting a landslide victory for Biden.

But, not so fast.  Ever since Mr. Trump’s surprise upset victory over Hillary Clinton in 2016 there has been considerable controversy over the reliability of presidential polling.  Pollsters have spent the greater part of the last four years analyzing why the polls failed to detect and report that the election was as close as it was.  Are they less accurate now compared to previously?  If so, why?  The answer seems to be that there are many theories, but no one really knows.

Some pollsters, such as Nate Silver of Five Thirty Eight, maintain that the polls are as accurate as they have ever been.  He maintains that they are correct 80% of the time.  Others, such as Jake Novak of Jake Novak News, pollster Glen Bolger and Matt Lackey, VP of Research and Development for Civis Analytics, whose opinions are outlined below, disagree.  In my view, since no one really knows how and why, the polls blew it in 2016 let alone how to correct them, it is likely to happen again.

I don’t want to drill down too far into the polling methodology.  That would bore all of you (and me) to death, but based on my research, I have some thoughts:

1. Novak and many others denote the obvious, which is that presidential elections are decided, not by the overall popular vote but by state-by-state electoral votes.  Therefore, the predicted winner of the overall national vote will not necessarily win the election.  Indeed this has happened five times, most recently in 2016 when Hilary Clinton lost to President Trump.  One can debate whether or not this is “fair” or “just,” but the rules are specified in the Constitution, and there are various reasons for them.  (Some of you may recall that I analyzed the role of the Electoral College in a previous blog.)  Can you name the four other presidents who were elected despite losing the popular vote?  See answers below.
2. Many pollsters query “registered” voters as opposed to “likely” voters.  The latter group would be more reliable, but they are also harder to identify.
3. As I said, national polls have proven to be reliable measurements of overall popularity, but they don’t necessarily predict the presidential winner accurately.  For example, in 2016 the final Rasmussen poll predicted Clinton to win the popular vote by 1.7%.  She won by 2.1%.  Not a bad prediction as far as it went, but it was misleading.  As we know, Clinton lost the Electoral College vote and, hence, the election.
4. According to Novak state polls have tended to be less reliable than national polls.   Lackey and pollster Glen Bolger agree and attribute that to (1) there is a segment of voters, most likely lesser educated and Trump supporters, that pollsters don’t reach, and (2) many Trump supporters are reluctant to admit it.  This view is supported by the American Association for Public Opinion Research, which opines that many state polls “underweight” Trump’s support because they often fail to “weight or adjust their samples to include more white voters who [haven’t] graduated college.”  Finally, according to Fox News a recently- published poll disclosed that 62% of voters are “reluctant” to express their true political preferences to pollsters, their employers, their friends or anyone else for fear of retribution (in the form of physical violence or loss of job) or of being mocked.  This feeling was expressed by Dems, GOPers and independents, alike, although it was more prevalent among GOPers.  Given the recent aggressiveness of the “cancel culture,” I find this very plausible.  I have to believe that the preponderance of these respondents are Trump supporters.

CONCLUSION

So, what does all this mean?  Well, for one thing, we should take the polls with a “grain of salt.”  I am not suggesting we ignore them, but we should be mindful of the lessons of 2016.  Mr. Trump’s claim that the polls are misleading and underweight his support may have some basis in fact.  According to Politico he “has a point.”

In my opinion, he may still be losing, but in all likelihood the election will be much closer than most people think.  If I were the Bidens I would not be measuring the White House for curtains just yet.

Quiz Answer:  (1) John Quincy Adams defeated Andrew Jackson in 1824.  (2) Rutherford B. Hayes defeated Samuel Tilden in 1876.  (3) Benjamin Harrison defeated Grover Cleveland in 1888.  (4) George W. Bush defeated Al Gore in 2000.