The Results of The Poll after The Presidential Elections in The USA in 2016

The 2016 presidential election between Donald Trump and Hillary Clinton was very confusing as the polls represented to the public did not match the outcome. Polls are supposed to provide an equal opportunity for Americans to express their opinions in a way that the election process cant. However, after this election, many people felt like they had gypped of their voice. Almost every political poll presented Clinton as the winner, but of course, this was not the case when Trump was elected president of the United States. Many wonder how the pollsters could have been so wrong, but there is no definite answer, only some theories. People can affect surveys greatly because “they can change their minds, they can decide to not share their opinions or they can flat-out lie (Chalabi, 2016).”The most talked about theory as to why the polls were wrong is non-response bias which is when an individual is unable or unwilling to answer the survey. According to www.pewreseach.org, “some groups – including the less educated voters who were a key demographic for Trump on Election Day – are consistently hard for pollsters to reach (Mercer, Deane, McGreeney, 2016).”

People with small levels of education and lower incomes are less likely to respond to any political survey either because they may choose not to if they are not educated on politics, or if they don't have a phone to be reached on. If there was a greater amount of uneducated or less fortunate people planning to vote for Trump than Clinton, then this would have caused the polls to be more in favor for Clinton. This is a very substantial sampling error that potentially affected the outcome of the polls during the 2016 election. The second theory as to why the polls were wrong is “The Shy Trumper Hypothesis.” This idea states that many people were sheepish to say they were voting for Trump since, at the time of the election, it was socially undesirable to vote for him. Most people had the desire to please (a sampling error) the pollsters so they may have told them they were going to vote for Clinton when in reality they were planning to vote for Trump. During this election, these people were known as “the silent majority” in which many people kept quiet about their intentions to vote for Trump in order to please the rest of society. This hypothesis is very closely related to a 1982 election in which “Democrat Tom Bradley, the black mayor of Los Angeles, lost the 1982 California gubernatorial election to Republican George Deukmejian despite having been ahead in the polls, supposedly because voters were reluctant to tell interviewers that they were not going to vote for a black candidate (Mercer, Deane, McGeeney, 2016).”

This theory of shy trump supporters could definitely be a reason for the polls saying that “Trump has a 15% chance of winning, roughly the same chance of rolling a total of six if you have two dice (Chalabi, 2016)” and underestimating Trump's level of support. The last theory is “the likely voter error” or the social desirability bias, in which is described as Americans saying they were going to vote but didn’t actually plan to do so. Some people may have planned to vote when they took the survey but changed their mind afterward. This change of heart could have been created by many peoples frustration with Trump's character and Clinton's potential criminal activity which left many people with no desire to vote after all.

I think in order to make a more accurate poll, pollsters should have created a late polling system. Since “FBI director James Comey's [came out with a] late decision to review additional Clinton emails (Boomey, 2016)” this could have swayed many voters at the last minute that many pollsters did not catch in their polls. The final average of the polls stated that “Hillary lead 45.5% to Trump's 42.2% (Boomey, 2016)” which could have been drastically different if there was a late polling system. It appeared as though Clinton “lead by a margin small enough that it might have been a polling error, (Bialik, Enten, 2016), in which many people may have assumed that since Clinton was in the lead, they didn’t need to go in and vote. I believe the election could have turned out differently if pollsters took another poll closer to election day, then more people planning to vote for Clinton would have gone in to vote if they knew Trump was winning.

To stay on the subject of politics and presidential elections I chose a poll from http://www.rasmussenreports.com/public_content/politics/trump_administration/prez_track_sep27 about the approval rate of Trumps presidency. The Rasmussen Reports states that they post new polls Monday through Friday at around 9:30 am. Since data is recorded every day for this survey, it is less likely the polls will be affected by a current event that would change people's opinion on presidential approval. The repetition of this survey allows for the polls to be more consistent with the trueness of the population's opinions on a day to day basis. On Thursday(09/26/18) the polls show that “47% of Likely U.S. Voters approve of President Trump’s job performance. 52% disapprove (Rasmusen, 2018).” Today(09/27/2018) however, 35% Strongly Approve of the way Trump is performing and 44% Strongly Disapprove. This gives him a Presidential Approval Index rating of -9 (Rasmusen, 2018).” In order to assemble the public's opinion, Rasmussen Reports uses telephone polling techniques in which they collect “telephone surveys of 500 likely voters (so a margin of error of 4.5%) per night and [report] on a three-day rolling average basis (Rasmussen, 2018).” They also eliminate some non-response bias by using an “online survey tool to interview randomly selected participants from a demographically diverse panel (Rasmussen, 2018).”

The subjects were contacted in two different ways of surveying, which would eliminate some people not being able to respond due to their lack of having a landline or cell phone. However, this leaves out people who may not have the technology to reply such as homeless people. Something that really stood out to me was the lack of information given in this article as to what (if any) background knowledge about Trump's presidency was given to the subjects at the time of the survey. Also, Rasmussen Reports says they contact “likely voters” instead of randomly selecting people from the public, which could create a bias of those who would be eager to respond. The article also not provide a response rate, so it is impossible to tell the percentage of the original subjects who actually provided information.

In conclusion, I think this poll has some great aspects by using a repetition method and different ways to contact the subjects, but there were key elements that this article left out to make this poll compelling. In order to make this poll more believable, the article should have provided; if the pollsters gave the subjects background knowledge on Trump's presidency, a random sampling method to pick the subjects, and a response rate. I think the graphs in the article made the poll more visually appealing, but for those who are statistically literate, this poll is not compelling enough to be believable.