• Fri. Mar 29th, 2024

The Official Student Paper of Riverside Poly High School

Presidential Prediction Disaster

Dec 5, 2016

UPSET: Shocking election results leave many popular polls incorrect, but confirm the predictions of several political scientists. 

By Jacob Ferrall, Staff Writer

Leading up to the 2016 presidential election, all major news sources and polls confidently expected a strong victory for Hillary Clinton. Popular news publications such as The New York Times predicted a drastic 85 percent to 15 percent Democratic victory, while established online polls such as Nate Silver’s “FiveThirtyEight” gave Trump a 33% chance of winning— a chance radically considered too high. These are just a few of the countless internet polls with inaccurate predictions. Current news media speculates a variety of reasons to explain why the polls missed the mark so much and how Trump managed to win the election.

Small scale presidential polls often cannot be trusted due to the limited sample of people and the potential bias of the group conducting it. A specific crowd is attracted to participating in polls, even more so when the surveys are conducted online. Generating a poll with only one source of information, also called a single poll, can also be very inaccurate, as biases are evident. A single or individual poll will only count the votes of a small number of people in a contained area. The citizens responding to this poll will predominantly be around the same economic standing due to these limitations. Unfortunately, most online polls are single polls, because they are much cheaper to create than a poll requiring extensive research. For a poll to have credibility, there must be more factors than citizen preference— which some political scientists accounted for.  

Political scientists from prestigious universities have maintained accurate predictions of elections for years. Ray Fair of Yale University predicted a Trump victory by 56% of the vote, and he was not the only political scientist speculating a Republican win. Stanford alumni Alan Abramowitz claimed Trump would win with 51.4% of the votes. These political scientists were very close to predicting Trump’s final result of 57% of the electoral votes. 

These scholars’ models take additional factors into consideration. Before making a final decision, they observe previous election patterns and familiarize themselves with the state of current voters. Fair explains that the data of the past twenty-five elections shows that “the economy has significant effects on voting behavior, as does the duration variable.” The duration variable is the estimated amount of time a political party will hold the presidency. Many mainstream polls did not consider that the presidency has shifted political parties about every two terms. Fair predicted the Republican candidate would receive extra votes due to the “sluggish economy and desire for change,” but he was still skeptical due to Donald Trump’s polarizing personality. 

Meanwhile, Abramowitz’ model accounts for the growing Gross Domestic Product (GDP). The GDP is the total value of all goods produced in the United States. In his model, the change of real GDP (adjusted for inflation and deflation) in the second quarter of the presidential election, as well as the current president’s approval rating, affects the outcome of the election. Obama only saw a small increase in his approval rating and there was a small increase in GDP, which was not enough to offset the challenge of the party holding office for three terms. 

However, even with a solid foundation, unpredictable events can occur. The reputable election forecasting site “PollyVote,” (with three correct previous presidential predictions) uses information from several different models as well as citizen forecasts, but still incorrectly projected the outcome. Post-election reporters say one wild card element of the election was the “silent majority” of Trump voters. This term, originally used during the 1968 election between Richard Nixon and Hubert Humphrey, describes the quiet supporters of a candidate who are drowned out by the harsher critics of the candidate. There were a large amount of rural Americans who came out on election day to give Trump their vote. 15% of Americans live in rural areas, so they most likely did not participate in any form of news media, but surprised the polls when they placed their votes for Trump. 

It may not have been very apparent during the election season, but there happened to be many voters who longed for a change in the government, who were not discouraged by the Republican candidate’s persona. During the presidential campaign season, Trump constantly alienated the U.S. Latino minority by asserting that he would deport all illegal immigrants and build a wall bordering Mexico. This led to the assumption of a minuscule vote for Trump from Latino votes, but Trump gained 29% of the Latino vote. This shocker may have occurred due to the fact that Latino citizens eligible to vote are registered citizens, who do not have to worry about deportation. 

The audience, source and content of a poll determine its final accuracy. There are a number of factors which many polls do not consider that act as turning points in the results. This 2016 presidential election has demonstrated that not all polls are to be trusted, and predictions of the outcome should be made with caution. 

Translate »