Click here to learn more about author Danette McGilvray.

If you’ve read my recent post about trusting election results (link to post) you’ve read my commentary regarding what I consider a data quality mishap during the US election of 2000. That exploration highlighted the necessity of high quality, trustworthy data to the election process.

Following the election, there were attempts to address the problems with tallying votes accurately—what I consider a data quality snafu. Efforts were made to eliminate the problem of the hanging chad. The emphasized solution was the adoption of new technology—electronic voting machines. By the election of 2004, the butterfly ballots were no more. The punch machines necessary for butterfly ballots had been decertified. Instead, electronic voting machines known as “D.R.E”s (direct recording electronic voting machines) were installed. While technology is a critical piece of the solution, it is not the only concern. A holistic data quality view of what could be done was not taken into account and the solutions proposed were too narrow.

In fact, as was discovered with the implementation of the D.R.Es, technology came with its own host of problems. Lately, there have been security concerns with the D.R.E’s. They have been vulnerable to hacking, which compromises the data. Additionally, many D.R.Es have proven to be sensitive to mechanical issues, which can slow the process and in some cases prevent voter participation. “People weren’t thinking about voting system security or all the additional challenges that come with electronic voting systems. Moving to electronic voting systems solved a lot of problems, but created a lot of new ones.”[i] –Lawrence Norden, The Brennan Center for Justice. In the end, attempting to patch what was fundamentally a data quality issue with a purely technological solution ended up creating its own headaches—and did not accomplish the end goal of producing trustworthy election results.

Recognizing the issue as a data quality problem opens up a holistic perspective that creates better solutions. The narrow focus on improving technology only ignored additional methods that may have been more effective. Training for polling attendants, for example, could help to implement a better process to ensure data quality. Better education of voters can help prevent some of the user errors that occurred in the election of 2000. Improving the technology that tallies the vote cannot guarantee that voters will understand that technology. Better designed business processes should not be foregone in favor of a costly and less effective solution.

It is impossible to know for certain which combination of solutions could have improved the past election data quality issues. However, it is certain that a data quality-inclusive solution is necessary to increase the accuracy and trust in election results in the future. History has shown us that zeroing in solely on technology blinds us to the processes and data quality methods that form a crucial part of the solution.

For those of us who work outside of the political world, this can serve as a lesson. Do you recognize the data quality component of problems in your organization? How do you approach those data quality problems? If your solution emphasizes technology to the exclusion of all else, you’ll need to take a broader view.

[i] Brian Barret, “America’s Electronic Voting Machines are Scarily Easy Targets,” in Wired. https://www.wired.com/2016/08/americas-voting-machines-arent-ready-election/.