How NEAR Accurately Predicted the 2020 US Elections Results Using Home Location Data

Back in October 2020, NEAR prophetically predicted the results of the 2020 US Elections by examining various factors like campaign strategies and change in home locations. By studying change in home location data, NEAR predicted that the key swing states would strongly lean towards the democrats as home location influxes in these states were mainly from other blue states. This prediction proved true for most of the states as seen in the election results. 

The Study

2020 was a very different year for the US presidential elections. From how the presidential campaigns were conducted to how the country ultimately voted, the Covid19 pandemic upended nearly every aspect of the election cycle. Rallies were held virtually, conventions were called off, debates were rescheduled, and a record number of ballots were sent by post. With the country’s caseload and death toll steadily rising and the economy witnessing an all-time low, the US elections witnessed major setbacks. NEAR followed the elections and the pandemic situation in the US very closely and prophetically predicted the results for nine out of the twelve swing states and the win for the Democrats. 

How did NEAR manage to make such a near-accurate prediction?

The team at NEAR studied economic displacement that occurred due to the pandemic and also conducted research based on Voter Sentiment. The research showed that there was a demographic layer to the displacement based on age and ethnicity. The study showed that the following important factors contributed to the results.


1) Demographic pattern showing displacement to swing states and its effect on the voting pattern 

NEAR studied Voter Sentiment (VS) based on three key factors: turn of national events from May 2020, campaign strategies, and change of home locations that affected election strategies. 

These factors helped NEAR give a highly accurate prediction of the US electoral results based on the data collected. First, we shall look into how the various national events in May 2020 resulted in significant displacements that ultimately contributed to the swing in votes.


2) Turn of national events from May 2020

Presidential campaigns and experts keep track of the shifting electoral landscape. Swing states for past elections were determined simply by looking at how close the vote was in each state. Deciding which states would be likely candidates as swing states in future elections requires estimation and projection based on previous election results, opinion polling, political trends, and developments that have occurred since the last election. The strengths or weaknesses of the particular candidate involved are also essential. 

NEAR conducted comparative research on how displacement to swing states from a Red (Republican) or Blue (Democrat) state during Covid 19 situation in May 2020 affected the electoral outcome. 

The swing-state “map” transforms between each election cycle, depending on the candidates and their policies, sometimes dramatically and sometimes subtly. The coronavirus pandemic saw an unprecedented amount of people impacted, and the death toll was three times higher than that of the Vietnam War. Compelling economic reasons and health factors resulted in people migrating from pro-Democratic or pro-Republican states to swing states. A swing state could be won by either a Democratic or a Republican candidate. These states are also referred to as Battleground states. NEAR used data to show that home location shifts due to Covid-19 in May 2020 was one of the primary reasons for the swing states to vote for the Democrats, contributing to their win.


3) Campaign Strategies

Often, presidential candidates campaign mostly in the swing states. Hence, a select group of states frequently receive a majority of the advertisements and candidate visits. Due to the displacement of people from pro-Democratic or pro-Republican to the swing states right before the election, NEAR’s research showed that the swing states’ campaign strategy had to be reworked. The pandemic complicated the campaigns ahead of the US election in November 2020. Campaigns shifted from in-person mode to virtual mode. Biden campaigned under pandemic response recommendations, hosting the majority of his messaging virtually. In contrast, Trump hosted in-person campaigning events, which were later classified as “super-spreader” events and ultimately resulted in negative publicity. NEAR’s research showed that the displacement of people also affected the gerrymandering strategies of both parties. It was difficult to predict which state remained pro-Democratic or pro-Republican or shifted to the other camp due to the migration from one state to another. 


4) States with the maximum displacement and its effect on the vote

Analysts came up with a list of a dozen swing states for the November 2020 US elections. They were Florida, Georgia, North Carolina, Arizona, Minnesota, Iowa, Wisconsin, Ohio, Michigan, Pennsylvania, Texas, and New Hampshire. Of these, six were believed to be crucial. The Cook Political Report identified Arizona, Florida, Michigan, Pennsylvania, Wisconsin, and North Carolina as most important in determining who wins the presidential election of November 2020. This would go on to significantly impact the election results. 

Let us look at why, with the below example set for the change of home location from California to Florida with an age-wise breakup. 

The above calculation indicates that the impact of displacement on the VS of Florida was 607.35 (total of the final counts) and that all those who migrated to Florida came from a pro-Democrat state. Hence, that increased the number of Democrat voters in Florida by that many digits. Similar tests were conducted for displacements from Texas (a pro-Republican state) to Florida (a swing state), to arrive at the Republican-leaning scoring for Florida. These studies were conducted for all the other states, and then the aggregate of the values was taken. Whichever party got the highest score in the swing state was identified and noted as the likely winner. 

Digging deeper, NEAR also found that the displacement was driven by age and ethnicity, which further impacted the votes.


5) Age as a factor for displacement and its effect on voting 

NEAR found that displacement to Florida based on age created a leaning towards the Democrats by 2.99 percent in Florida and 23.39 percent in North Carolina. Wisconsin, Pennsylvania, and Michigan, all crucial swing states, also showed a strong leaning towards the Democrats. Those displaced were mainly between the ages 18-44, as seen from Figure 1 above. Hence, NEAR concluded from its data collected that the displaced population favored the democrats, and the younger generation had a strong leaning towards the Democrats. Another crucial factor was that it was the ethnic minority group that experienced displacement the most and that impacted the result.


6) Ethnicity as a factor for displacement and its reflection on voting

Minority groups such as Black, Hispanic, and Native American communities were the worst hit by the pandemic due to decades of socioeconomic factors (based on demographics data on an aggregated level). These communities are primarily clustered in high-density urban localities and were most affected by the pandemic in the first few months.

NEAR found that these communities that got displaced to the swing states all favored Democrats to a large percentage. The six states that were stated crucial by the Cook Political report showed a clean leaning towards the Democrats. NEAR’s research showed that the six crucial states were usually Red bastions but got converted to Blue due to displacement that swung the results in favor of the Democrats.

NEAR’s research also showed that factors like unemployment, economic downturn, and deteriorating health conditions in crowded urban areas were a significant disruptor. It led to large-scale migration from one state to another, and that impacted the election results. 

Near’s Accurate Prediction

NEAR's prediction that the swing states would majorly vote for the Democrats, as indicated by the study, was seen to be accurate for nine states out of the twelve swing states and the Democrats did go on to win the election. The aggregated and anonymized data on home locations and the shifts in them due to displacement played a major role in enabling NEAR to arrive at the correct conclusion.