Exit polls are one of the most important tools for identifying supporters of a specific candidate or party in U.S. elections. Various media outlets station people outside sample voting stations to poll voters on how they voted immediately after they exit the station. The results of these polls are usually based on huge samples taken from across the country and they are fascinating.
Midterm exit polls showed that around half of voters said they had decided who to vote for "in the last month" and even "in recent days." In this respect, there was no difference between supporters of Democratic and Republican candidates for the House of Representatives. The polls indicated the final results long before the election was held. The same was also true in the past. When U.S. President Donald Trump won the 2016 presidential elections, many commentators called it a "big upset," but it turns out early exit polls had been right.
In the congressional elections, the exit polls were right about 47 candidates across all 50 states and the District of Columbia. In places where they erred, the polls had pointed to a tight race. Three of these states, Florida, Pennsylvania and Michigan, have a substantial number of electoral votes and they all tipped the scales in U.S. President Donald Trump's favor. The exit polls predicted that then-Democratic presidential candidate Hillary Clinton would win the popular vote by a margin of 2% and that is in fact what happened.
Many believe that voters' positions on political and social issue determine how they vote, but two trends indicate this is not necessarily the case. First, supporters of the ruling party believe their personal situation and the state of the nation are better than those who support the opposition. In a CNN exit poll, 86% of those who voted Republican said they thought the U.S. was heading in the right direction, while 85% who voted Democratic said it was not. Second, voters were familiar with the ideologies and policies of each party. Of those who described themselves as "liberal," 91% voted for Democrats, while 94% of those who described themselves as "conservative" voted for Republicans. Ninety percent of those who said Trump's immigration policy was too strict voted for Democrats, while 85% of those who said it wasn't tough enough supported Republicans.
The best predictor of how someone will vote is party affiliation. Around 95% of those who identify as either Republican or Democrat supported their party's candidate on Election Day. Ninety-four percent of those who supported Clinton in the 2016 presidential elections supported a Democratic candidate in the 2018 midterms. Of those who voted for Trump in 2016, 91% voted for a Republican candidate in 2018. This strong party loyalty aligns with a conclusion reached by University of Michigan researchers 60 years ago: Party identity dictates voter stance on any given issue more than the stance on an issue dictates party affiliation. A similar phenomenon exists in Israel.
Exit polls in elections in past decades have shown similar findings. The importance of party loyalty then begs the question: What causes a political upset? Is it demographic changes to the electorate or other factors that change someone's party affiliation over the years? In fact, it is not the voters who belong to either of the two largest political camps who are responsible for the dramatic upsets we see. It turns out that around one-third of U.S. voters do not identify as either Republican or Democrat, or as conservative or liberal. Party affiliation is a decisive factor in voter conduct, and the independents hold the key to revolution.
