Elections 2020: why people vote, how people vote, and what lessons we learned from 2016 polls.

Tomorrow is the most important day of the year for the United States. Here at Narrative BI, we decided to analyze tons of open-source data to give you a perspective on the 2020 elections, how the 2016 elections experience can help us predict results, and what motivates people to vote in 2020. Let’s find out!

Candidates perception difference

We took the data from Ipsos’ KnowledgePanel and defined several topics people care about. Some of these topics are more important for Trump supporters: for example, “society becomes too soft and feminine” (59 pp. more for Trump), “police actions are appropriate against vandalism” (40 pp. more for Trump), “the country needs a strong leader” (24 pp. more for Trump) and “military methods are efficient in governing the country” (24 pp. more for Trump).

Biden supporters are more concerned about racism in the police (73 pp. more for Biden). They are also confident in different sources of information like media (70 pp. more for Biden), the Center for disease control (51 pp. for Biden), and they trust electronic vote (55 pp. more for Biden).

According to Pew Research Center data, it’s essential how candidates will handle BLM and racism topics this year because African Americans demonstrated a significant increase in turnout rate for the last four election campaigns. The only exception is the 2016 campaign.

Turnout rate, 2004–2016

Interesting fact: democrats are more loyal to their candidates than republicans. Among Democrats, there are only 2.6% who are not going to vote for Biden, and among republicans — there are 5.2% such voters.

By the end of October, more than 76 million Americans already made their decision. We have more data in the states that have party registration, and we can find some patterns in COVID-19 affected voting (we will talk about in a second).

As you can see, Republicans are voting more often in person. They’re less active in using mail ballots (both returned and outstanding). We see this as a critical difference in voting patterns between parties.

Some groups are going to vote more than others: who had a friend or family member test positive COVID-19 (5% more), who think that the government should make decisions, not expert groups (5% more), people with income more than 125K a year (6% more). Religious people, unemployed for more than a year, and people with complicated financial situations tend to pay less attention to elections this year.

Voting behavior factors

2016 vs. 2020 polls

The majority of polls are showing that Biden has strong chances to win. According to FiveThirtyEight, Biden right now is 8.9 percent point ahead, but we learned that polls are not the definitive source of truth: Hillary Clinton has a 12-point lead over Donald Trump in 2016 (50% against Trump’s 38%).

We want to talk more about the 2016 polls and what we can learn from experience. There was a group of states with significantly inaccurate polls during the last elections: South Dakota (+16.3 % to the final result), West Virginia (+15.7 %), Oklahoma (+14.8 %), Wyoming (+14.5 %), and Kentucky (+14.1 %).

New York (+5.1 % to the final result), California (+5 %), Nevada (4.1 %), Hawaii (+2.8 %), and California (+2.1 %) had the most accurate polls.

2016 results — Polls difference in %

It seems that is the most unstable situation is in traditionally republican/democratic states. The result can be very unpredictable. Let’s check where Trump and Biden have support according to the 2020 presidential polls.

Trump has the best polls in Wyoming (67.6%), West Virginia (63.1%), Oklahoma (61.7%), Idaho (60%), Arkansas (59.6%), North Dakota (59.6%), Alabama (59.4%), Kentucky (58.9%), Louisiana (58.5%), Nebraska (57.9%), Tennessee (57.4%), South Dakota (57.1%), Mississippi (56.6%), Utah (56%), Kansas (54.5%), Indiana (54.2%), Missouri (53.9%), South Carolina (53.8%), Montana (52.9%), Alaska (52.8%)

Biden has the best polls in District of Columbia (90%), Massachusetts and Vermont (66.6%), Hawaii (66.4%), Maryland (64.2%), New York (64%), California (63.75%), Rhode Island (63.2%), Delaware (61.4%), Connecticut(61%), Washington (60.5%), New Jersey (59.2%), Oregon and Illinois (58%), Virginia (55.7%), New Mexico (55.4%), Maine and Colorado (55.2%), New Hampshire (54.7%), Wisconsin (53.7%).

There are some unstable states where polls fluctuate more: Arkansas (+-5.2%), Hawaii (+-3.9%), West Virginia (+-3.6%), Massachusetts (+-2.7%), Maine (+-2.8%). Biden improved his positions in all states above, except Hawaii where Trump’s polls became better.

Let’s dig deeper into the state voting stability. We discovered that some states have a significant probability that an individual voter in the state will determine the Electoral College winner.

There is 15% bigger probability compared to the average that an individual voter will decide who wins the 2020 elections in North Carolina, New Mexico (+16.7%), Michigan (+21.3%), Arizona (+27.4%), Minnesota (+29.6%), Florida (+30%), New Hampshire ( +33.3%), Nevada (+40.4%), Wisconsin (+52.2%), Pennsylvania (+57.8%).

These states are critical: they can bring the Electoral College’s decisive vote even if the national vote shares are close.

State voting stability map: green-stable than average, white-average stability, red-less stable

We hope you liked our analysis! Feel free to subscribe to our LinkedIn and Medium page to get the best analytics on hot topics 🙂 This week we will also present the second part of our big COVID report, where we give you deep insights and predictions on economic situations in different US states. Stay tuned and jump in the comment section to share your thoughts on the 2020 election!

A new user-friendly tool that brings meaningful data insights to the user without a single click. Check out www.narrative.bi for more!

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store