How we cast votes in federal elections is broken. We elect representatives into offices with terms that may last as many as six years, yet the process for casting a vote must be done within a 12-hour window, on the first Tuesday after the first Monday in November. The roots of a designated day for voting dates back to 1845, when congress declared a national election day. There is much at stake to get every and all votes verified and counted, to reflect the preferences of the majority.
Regrettably, voter turnout for presidential elections is dismal, with midterm elections even worse. This makes each election not an indicator of population sentiment, but rather, a sample of this sentiment. Since the presidential election of 1932, the highest proportion of the voting age population that cast a vote in the presidential election was nearly 63 percent (in 1960), while the lowest proportion was 49 percent (in 1996). This means that between one-half and five-eighths of eligible voters determine who wins the most important and powerful office in the world.
Between eight and 12 battleground states determine who wins the White House. Electoral college votes in these states are earned by which party can get the majority of voters to show up. Placing the outcome of such a high stakes election on the uncertainty of a single day is neither reasonable nor sensible. Adding COVID-19 into the uncertainty equation this year makes this antiquated ballot casting process even more risky and unacceptable. Given that much has changed in our nation over the past 175 years since the national election day decision was enacted, perhaps it is time to rethink how we cast votes for those who represent our interests, spend our tax dollars and craft our national and foreign policies.
Adding both technology and time to the election process can improve the likelihood that the sample of votes cast reflects the sentiment of the population.
Our internet-of-everything, artificial intelligence driven world has transformed numerous facets of everyday life. We can access our bank account on the other side of the globe with a piece of plastic containing a microchip and a four-digit pin. The census bureau relies on the internet to collect demographic information that has far-reaching economic and social implications over the next decade. We are placing an ever-growing reliance on computer systems and machine learning to diagnose diseases and improve population health. Yet, we view with suspicion the possibility of using similar technological advances to choose our representatives in state and federal elections. Why can we not accomplish the same level of trust, verification and anonymity of votes with our elections that we assume in so many other places?
Security breaches occur and make splashy headlines. Yet, when we consider how much of our information is available, and the small percentage that gets used maliciously, we realize that precautions can be taken to ensure that the results from a well defined event — like an election — can be secured. Security breaches are a red herring in the pursuit and implementation of technology for casting ballots.
Even retaining paper ballots, as the AAAS EPI Center recommends, does not preclude technology from facilitating the collection, counting and validation of ballots. Adding multiple methods for casting ballots is sensible, it empowers voters and increases the likelihood of higher voter turnout.
One unspoken reason that technology is often rejected by politicians to manage elections has nothing to do with security. In almost every election, one candidate has a voter preference edge over another candidate. Technology would make it more difficult for the underdog candidate to overcome any deficit if every eligible voter is able to cast his or her vote. In such cases, the sample preference approaches the population preference, which minimizes the impact of politics on the election process. This means that politicians and political parties are motivated to retain the uncertainty of who will vote, effectively providing a lever to manipulate election outcomes. The book, “Victory Lab” openly describes this process.
The second factor is time. With technology in place to facilitate voting and the casting of ballots, a voting day can become a voting week or even a voting month, as was the case in the early 19th century. Widening the window of opportunities to cast ballots can only increase the percentage of eligible votes cast. Perhaps mobile or virtual voting centers can be launched to bring the opportunity to cast a vote to the voters, rather than asking the voters to visit voting precincts.
There is no one size-fits-all solution. Multiple methods by which ballots can be cast widens the footprint of opportunities for voters to exercise their constitutional right to vote. Simplifying ballots will also make it easier for voters to express their choices. Perhaps even decoupling federal, state and local elections so that not all elections are held at the same time, with the same ballot, to further simplify the voting process.
Elections are about political choices, yet there is no place for politics in the election process. Just as gerrymandering tarnishes how congressional districts are drawn, casting ballots using antiquated and ineffectual rules and regulations taint election processes. The time is ripe to get politics out of political election voting processes, and allow the voters to regain the power of choice that our democracy affords them.
Sheldon H. Jacobson, Ph.D., is a founder professor of computer science at the University of Illinois at Urbana-Champaign. He is the founder of Election Analytics, a STEM learning laboratory focused on forecasting the outcome of the United States Presidential election, and the Institute for Computational Redistricting (ICOR), a research program for advancing algorithmic redistricting to combat gerrymandering.