Wednesday, September 28, 2016

Why You Shouldn't Believe Poll Numbers

Why You Shouldn't Believe Poll Numbers
by Angela Smith, HEAL Coordinator


The Commission on Presidential Debates requires that candidates receive 15% in five different national polls in order to qualify.  The polls they are relying on in the 2016 election are: ABC/Washington Post, CBS/New York Times, CNN/Opinion Research Corporation, Fox News, and NBC/Wall Street Journal.(1)  ABC/Washington Post based their poll results on a total of 1002 respondents of which only 647 were "likely voters" in the 2016 election.(2)  CBS/New York Times based their poll results on a total of 1753 respondents of which only 1433 were registered or likely voters.(3)  CNN/Opinion Research Corporation based their poll results on a total of 1001 respondents of which only 886 were registered or likely voters.(4)  Fox News based their poll results on a total of 1006 respondents, all of whom were registered voters, and 867 which responded they would likely vote in November.(5)  NBC/Wall Street Journal based their poll results on a total of 1000 respondents, all of whom were registered voters.(6)  That means a total of 5762 people were polled and are deciding who is permitted to participate in the 2016 presidential debates.  That's a lot of power for those 5762 respondents and seemingly undemocratic. 

There were 129,100,000 voters in the 2012 general election.(7)  And, 5762 is .00446% of all voters.  "The larger your sample size, the more sure you can be that their answers truly reflect the population."(8)  The average of the 5 poll numbers for Green Party Candidate Jill Stein provided by the outlets relied upon by the Commission on Presidential Debates is based on the response of 5762 people total and averages 3.2% and the poll questions are skewed to favor the two major parties.(9)  A local Pennsylvania Fox News poll that sampled over 77,000 respondents (over 10 times the aggregate of the 5 "official polls") showed Jill Stein winning in a landslide with 85% of the vote.(10)  An NBC poll that sampled 69,823 respondents (again, over 10 times the aggregate of the 5 "official polls") showed Jill Stein coming in second to Trump with 19% of the vote.(11)  Based on a rudimentary understanding of statistics, it would seem that the polls with ten times more respondents would provide a more accurate result.

Now, let's just say all the polls are unreliable because they do not provide a sample size effective to determine the true position of most voters in the United States.  When you look at the Brexit debacle in the United Kingdom versus the polls leading up to that vote, you can see polls really are not valid indicators regarding voter sentiment.(12)  If we take our total referenced polls in this article, which is 7, and say 2 show Jill Stein winning or coming in 2nd, then you have 2 out of 7 with any arguable accuracy.  2/7 = 1/3.5 which is close to the 1/3 number of polls whose numbers correlated with the actual result of the Brexit vote in the United Kingdom.  But, there are flaws with this reasoning, so let's agree to throw out all polls.

Now that we have no polls, how do we determine who is likely to win in the 2016 general election?  Well, let's take a look at Facebook "likes" of all four candidates.  We could consider Twitter "follows" and Facebook "likes", but, many people "follow" people on Twitter for reasons that have nothing to do with supporting the people "followed".  So, examining Facebook "likes" may provide a better basis for making any determinations.

On the official campaign pages on Facebook, the breakdown on September 25th, 2016 is: Jill Stein has 576,617 "likes", Gary Johnson has 1,534,003 "likes", Donald Trump has 10,800,664 "likes", and Hillary Clinton has 6,156,232 "likes".   If we base our "poll results" solely on Facebook official "likes", we work from the aggregate total of 19,067,516 "likes".  Based on Facebook likes out of all possible likes, we have Stein at 3.02%, Johnson at 8.04%, Trump at 56.46%, and Clinton at 32.28%.  Calculating with the Facebook numbers alone, it would only represent 14.76% of voters and only if all "likes" were doled out "one per customer" and represented actual or "likely" voters in the 2016 election.  The total number of users of Facebook in the US is 162,900,000.  7,500,000 Facebook users are 13 or under.(13)  There are 25,000,000 teens aged 12 to 17 in US.(14)  And, 73% (or 18,250,000) of 12 to 17 year olds use Facebook.(15)  So, we can subtract 25,750,000 from the total number of Facebook users in the US to get the number of voting age users (18 and older) and that total is 137,150,000 which is more voting age users of Facebook than people who voted in the 2012 general election.  In 2012, only 7.3% of Facebook users reported "voting".(16)    7.3%  could be read as 7.3% of all voters given that more US adults are on Facebook than voted in 2012.  And, I'm sure you can see why even this gets tricky and wouldn't result in truly comprehensive and accurate statistics or poll results.  But, it seems on its face more plausible than the polls we are supposed to "respect" and rely on.  But, is it really?  Or, should we ignore these statistics as well?

The fast food chain McDonald's has 67,160,654 Facebook "likes".(17)  The fast food chain Burger King has 7,905,567 "likes".(18)  This makes the total "likes" of both chains on any given day 75,066,221.  And, McDonald's receives 89.5% of those total "likes" with Burger King receiving 10.5% of those total "likes".  Based on these numbers, one might assume that McDonald's is 89.5% more likely to be the restaurant of choice of most fast food consumers when choosing between Burger King and McDonald's.  McDonald's has 36,525 locations world wide.(19)  McDonald's serves 68 million customers per day.(20)  Burger King had between 13,667 in 2013 and now reportedly has over 15,000 locations worldwide.(21)  Burger King claims 11 million customers per day.(22)  Now, based on customers served alone, it shows that about 86% of consumers prefer McDonald's to Burger King.  But, Burger King has approximately 41% of the number of locations when compared to McDonald's.  And, based on the numbers it would appear each Burger King location serves an average of 733 people per day.  Each McDonald's location serves an average of 1861 people per day.  If you add the total number of people served each day at any given location at both restaurants, you get 2594 total customers served per day.  Based on a side by side comparison of consumers per restaurant per day, Burger King receives roughly 28% of the total customers per restaurant per day and McDonald's receives 72% of the total customers per restaurant per day.  So, the Facebook "likes" do not provide an accurate picture of the true favorability based on consumer behavior when analyzed through existing factual data.  In fact, the facts show that Burger King is nearly three times as favorable, receiving 28% of the consumer support than is reflected in the Facebook "likes" analysis. 

What if we looked at each of the candidates based solely on small individual donor contributions to their presidential campaigns?  Would that give a more accurate idea of the voter support for each of the four 2016 presidential candidates?  Hillary Clinton has received $70,714,091 (19% of her total contributions) in small individual donor contributions.(23)  Donald Trump has received $48,353,930 (29% of his total contributions) in small individual donor contributions.(24)  Gary Johnson has received $5,534,943 (70% of his total contributions) in small individual donor contributions.(25)  And, Jill Stein has received $1,134,508 (60% of her total contributions) in small individual donor contributions.(26)  This gives us an aggregate total, of all small individual donor contributions to all four candidates, of $125,737,472.  This gives Hillary Clinton roughly 56% of the individual donor financial support.  This gives Donald Trump roughly 38% of the individual donor financial support.  This gives Gary Johnson roughly 4% of the individual donor financial support.  And, it gives Jill Stein roughly 1-2% of the individual donor financial support.  Do these results reflect election results?

In the 2012 general election, President Barack Obama received 65,918,507  (51.01%) total votes and Republican Nominee Mitt Romney received 60,934,407 (47.15%) total votes.  Gary Johnson received 1,275,923 (0.99%) total votes.  And, Jill Stein received 469,015 (0.36%) total votes.(27)  Now, what did their political contributions look like in 2012 in small donor contributions? 

President Barack Obama raised $234,388,190 in small individual donor contributions in 2012.  Mitt Romney received $80,058,900 in small individual contributions in the 2012 election.(28)  Gary Johnson raised a total of $780,345 in small individual contributions in the 2012 election.(29)  And, Jill Stein raised a total of $319,221 in small individual contributions in the 2012 election.(30)   The aggregate total of all donations to all four candidates in 2012 was $315,546,656.  This gave President Obama roughly 74% of the small individual donor contributions total.  This gave Mitt Romney roughly 25% of the small individual donor contributions total.  This gave Gary Johnson roughly 0.25% of the small individual donor contributions total.  And, this gave Jill Stein roughly 0.10% of the small individual donor contributions total.  But, as shown above, President Obama received 51.01% of the vote, not 74% as the donations to his campaign might otherwise suggest.  Mitt Romney received 47.15% of the vote, not the 25% donations to his campaign would suggest.  Gary Johnson received 0.99% of the vote, not 0.25% which donations to his campaign would suggest.  And, Jill Stein received 0.36% of the vote, not the 0.10% that donations to her campaign would suggest. 

Based on the above, we have Clinton receiving about 1/3 of the donations that President Barack Obama received in the 2012 election cycle, and he was an incumbent.  And, Donald Trump receiving about 2/3 of the donations that Mitt Romney received in the 2012 election.  Gary Johnson has received roughly 700% more donations than he received in 2012.  And, Jill Stein has received roughly 355% more donations than she received in 2012.  While the Democrats and Republicans have lost between 1/3 and 2/3s of their overall support, the third parties have grown their support exponentially.  Of course, this is based on current numbers available in the 2016 election and these numbers may change throughout the election cycle.

What does any of this mean?  Well, even an intelligent person trying to find the true and accurate percentage of voters supporting any individual candidate based on available objective data will find predicting the outcome of any election very difficult and the true percentage of voters backing any particular candidate virtually impossible until election day. 

As voters and citizens, we are responsible for electing government officials who represent us domestically and globally.  For the majority of us, Hillary Clinton and Donald Trump do not represent our domestic nor international interests.  We should demand a working democracy that includes all candidates, who are on enough ballots to win the election, be included in any scheduled general election debates.  We can achieve that by getting independent and third parties to 5% (or more) of the votes in the general election.  The polls are biased and rigged.  So, our only choice to save democracy is to vote third-party and ignore the lies our media tells us about what our choices are and whether we have the power to change that.  We do! 


Citations:
































No comments: