Why Were 2016's Polls so Inaccurate?

Earlier this fall, most polls had predicted there being at least a 70% chance of Hillary Clinton winning the presidential election. Despite winning the popular vote, Hillary Clinton failed to gain the majority of votes from the Electoral College. Discrepancies between predictions and actual results are likely due to underestimates about turnout of certain groups, the preferences of those groups, and unrepresentative samples. It also might be plausible to explore the effect voter ID laws, specifically in States like Wisconsin, to determine if such laws had an influence on voter turnout.Perhaps the biggest reason the results seemed to be so unexpected was the underestimation of how many rural white voters would turn out. There was a higher turnout of white voters without a college education than previously anticipated. At the same time, a larger proportion of the votes of white women went to Trump. This could have been a result of white women saying they were going to vote for Hillary in polls while in reality they were “secret Trump voters”. Poll predictions believed that many white people who voted for Obama would follow suit and vote for the democratic candidate in this election without any regards to the preferences of the voters. Trump is disliked by elites, media, corporations, and other politicians. While in the opinions of many across the country he is disliked for a good reason, it gave others, especially in rural areas, incentive to vote for him. Many believe that politicians are only self-interested and disregard the needs of voters. They were content that Trump had come from a background that lacked politics and government involvement. While many of his voters may have “nothing to lose”, it also highlights ignorance from these people in regards to systemic issues involving race, gender, religion, sexual orientation.In states with strict voter ID laws, like Wisconsin, it is also possible that laws had an impact on how many people were able to vote. The Milwaukee Journal Sentinel reported a decrease of over 40,000 voters from the election in 2012. Many of these 40,000 were in districts with higher levels of poverty that make it more likely to have problems with the voter ID requirements. In turn, if those 40,000 voters from the most diverse place in Wisconsin did end up voting, the results for the state may have been different. However, it has been argued that the reason for the decline of voters may be because of the unpopularity of both Trump and Clinton and not because of voter ID laws. Polls seemed to overestimate the amount of people of color who were going to vote for Hillary. Again, many polls based their predictions off of the 2012 election without considering there might be a higher number of votes from ethnic minorities for a black man than there would be for a white woman.If there is anything to be learned from this election, it is that it is impossible to predict what is going to happen in a social context. It is important to regard the information missed in poll predictions this time around to more accurately and efficiently predict information in the next election.