The 2020 Presidential Election: Impacts of Voting-By-Mail

Over the course of the last two presidential elections (2012 and 2016) and the 2018 mid-term election, most voters are polarized within the two-party system (Republicans in red and Democrats in blue), that represents the political environment of the United States. In the U.S., the presidential general election is not decided by popular vote, but by a process known as the electoral college. This process allocates each state with an electoral college vote commensurate with the total number of representatives in the Congress, one vote for each member of the House of Representatives, and one vote for each Senator. The capital of the Unites States, Washington D.C., is allocated a total of three electoral college votes, as outlined in the U.S. Constitution.

The map of counties shows a deeply divided political atmosphere, and at first glance, an overwhelming majority the country is represented by the republican party. However, population density and the percentage of the population that represents each party can often be overlooked in such a visualization.

The tally of the popular vote in the U.S. is done by county. The map above shows the current county voting preference by party. Counties that have not changed political party preference are shown by either the dark red or dark blue. For counties that have changed preference, the lighter shade of blue or red applies. Counties whose preference is likely to swing to either party is annotated in purple. There were no counties in the country that voted with a score of (0,1,0).

In the aftermath of the Coronavirus pandemic, there will be many changes within our society. With the upcoming presidential election, there is already speculation as to how politics will be conducted from now on. Implementing a national Vote-By-Mail (VBM) policy is likely to be a reality in future elections.

Currently, five states conduct all elections entirely by mail.(3) Utah fully implemented a mail in ballot system in 2018, after a statistically significant report was published outlining the effect that a VBM system impacted voter turnout in the 2016 presidential election.(4)

Voting trends from the 2018 mid-term election show that the states in purple are swing states for 2020. This is based on no changes in how the nation has previously cast ballots for the presidential general election.

Using the metrics determined in the Utah voter file analysis, the map below shows predicted outcomes if VBM statistics hold true for the rest of the United States. Specifically adjusting by political party and race, the map shows another side to voter party preference in the United States.

Voting-By-Mail creates a new political map, if properly implemented, with national implications for voter preference. Voter preference for the Democratic Party changes from 12 states to 35 states, while voter preference for the Republican Party drops from 20 states to 13 states.

No states remain designated as swing states. Seven states actually turn from "red" to "blue" with the increased participation of voters due to implementing a VBM system nationwide.

Unfortunately, getting more people represented in the voting process with a VBM system is easier said than done. Not all states are equal in their approach to how such a system should work. Currently, 17 states have some form of VBM system in place, with California, Nebraska and North Dakota allowing counties to determine for themselves how to enact VBM (with some prerequisites).

For the states that currently have a VBM system, there is no common process. For Wisconsin, first time voters are required to mail in a copy of their state identification (student identification is invalid). Additionally, along with six other states, a witness is required to verify the ballot with a signature. Alabama requires two witness signatures, but will accept a notarized ballot in lieu of signatures. Meanwhile, Missouri only accepts mail in ballots stamped by a notary public. A missing signature is cause for a ballot to be not counted, and the same is true for illegible signatures, or signatures that don't match the voting registration signature kept on file with the state. In extreme cases, reports of states not accepting signatures outside the signature box were counted as a reason for an invalid ballot.

Problems with VBM systems don't stop with signatures. Reportedly, 4%-20% of U.S. mail is either lost, or delayed in transit, which could be a problem for implementing a timely and fair VBM system. Also, a VBM system would render same day registration initiates futile, at least without an online or remote alternative.(5)

Overall, how VBM initiatives are implemented is just as important as why they should be implemented.

Political party voting data was gathered exclusively from Dave Leip’s Atlas of U.S. Presidential Elections and verified against voter file records compiled for each state by the United States Elections Project. Shapefiles were sourced from the U.S. Census database, as well as the race and population data. Citizen Voting Age Population data was sourced from the 2014-2018 5-Year American Community Survey, compiled by the U.S. Census.

All data was compiled at the county (administrative boundary 2) level and collapsed into the state (administrative boundary 1) level. Data was also compiled into the political boundary, by Congressional District, however this level only applies to Nebraska and Maine and became irrelevant after the results of what the VBM effect could have on election outcomes.

The data compiled by the VBM report in Utah was basis for this project. VBM projections are based on the statistical modeling in the report Utah 2016: Evidence for the Positive Turnout Effects of ‘Vote At Home’ (Also Known as Vote By Mail) in Participating Counties. Relevant voting populations were projected to Vote-By-Mail by both party affiliation and race. Republican VBM turnout was increased by 5.3% from the total votes cast in 2012, while democrat VBM turnout was increased by 7.4%. Race turnout was increased based on the percent of the population reported to identify as either Asian (+8.6%), Black (+9.8%), Hispanic (+6.1%), Other (+8.7%) and White (+6.9%), compared against the voting age population of the county by race. This population was then increased by the difference that VMB had on that race and applied against the voter turnout population of 2016.

Swing states were designated as states where neither party held a majority with greater than or equal to 50% of the predominant party vote. All parties not designated as Democrat or Republican were designated as a single "Other" category.