In our era, with money and ideas traveling quickly over the internet, the old saw that “all politics is local†may be less true than it used to be. But the administration of elections in the United States remains doggedly local. As a result, election results are published by state and local agencies in idiosyncratic formats. Any study of election results at a national level, or even across several states, requires painstaking assembly of disparate formats. This project builds a system to efficiently assemble election result data sets from several states into a single format. In addition, the project implements a variety of algorithms for fast, broad detection and visualization of anomalies in election results. The system will quickly flag significant anomalies in preliminary election result data from the general election in November 2020. This analysis will be made available publicly and to candidates, parties and election administrators. Candidates, parties and election administrators have the detailed local knowledge required to assess whether any particular anomaly in their district has a legitimate explanation or whether further investigation is appropriate.
In the best of times, conducting a trustworthy election poses many logistical challenges. In 2020, in addition to recent changes in election technology and widely acknowledged threats of cyberattack, there is the unprecedented complication of a raging pandemic in a Presidential election year. The project will provide a significant tool in the election verification toolbox by helping to identify anomalies well before election results are finally certified. Candidates and other interested parties will be able to use this analysis to inform timely decisions about whether to challenge or demand investigations into election results – before the determination of winners is set in stone.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
