It’s Friday morning, and by this time I think that everyone who follows American elections thought that we’d have some clear sense of the outcome of the Iowa Democratic Caucus.

Instead, we have headlines like this, from the New York Times, “Iowa Caucus Results Riddled With Errors and Inconsistencies.” While it’s not necessarily surprising that there are errors and inconsistencies in the current tabulation reports from the Iowa Democratic caucuses, the issue is that we may never get a clear, trustworthy, and accurate tabulation of the caucus results.

It’s helpful that the caucuses produced tabulation results on paper — and these paper tabulation records can be examined, and these records can form the basis for recounting and even auditing the caucus results. But it doesn’t seem that there ever was any intention for anyone to try to audit or validate the results of the caucus. And I keep scratching my head, wondering why, given how close and competitive the Democratic presidential selection contest has been, it doesn’t appear that anyone considered building a process to audit and validate the caucus results in near-real time.

For example, in our Monitoring The Election project, we pilot tested independent and near real-time quantitative auditing of a number of aspects of the election process in Orange County (CA) in 2018. We are now just starting to do that same type of auditing in both Orange County and Los Angeles County for the March Super Tuesday primary (we’ll start releasing some of our auditing reports very soon). A similar process could have been used in the Iowa caucuses.

What would it involve? Quite simply, the Iowa Democratic Party could work out a data provision plan with an independent auditing group (say the Caltech/MIT Voting Technology Project and/or university and college teams in Iowa). They could securely provide encrypted images of the tabulation reports from the caucus sites, and the independent auditing team would then produce auditing reports for each round of tabulation. These reports, like those that we currently produce as part of our project, would of course be provided to the appropriate officials and then posted to a public website. As rounds of tabulation proceed, this process could continue, until the final tabulation is complete, at which time the independent auditing group could provide their evaluation of the final reported tabulation.

This could have been done earlier this week, and had such a system been in place, it might have helped provide an independent perspective on the problems with the initial tabulations on Monday night, and quite likely could have alleviated a lot of the rumors and misinformation about why the tabulation was proceeding so slowly and why the results were riddled with errors and inconsistencies. By announcing, in advance of the caucuses, a plan for independent auditing of the tabulation results by a trustworthy third-party, the Iowa Democratic Party could have relied on the auditing process to help them figure out the issues in the tabulation, and perhaps helped to buttress confidence in the accuracy of the reported results.

At this point in time, while the data is being released, it’s unfortunate that there wasn’t an independent auditing process established before the crisis hit.

In my opinion, one of the most important lessons from this experience this week is that election processes need to be fully and independently audited. Whether those audits are conducted by academic researchers, or by other third-parties, they need to be a regular component of the administration of any public election process (caucuses, primaries, special elections, and general elections). I think that election officials throughout the United States can learn a lesson about the importance of independent election performance auditing from the chaos of the Iowa Democratic caucuses.