What the failure of the Iowa Caucus mobile app can teach us about digitising democracy
This year, the Iowa caucus was set to make history by being one of the first caucuses or primaries in the United States where the votes were recorded and submitted via a mobile app. The Democratic Party hoped that the use of an app would make counting and reporting the results of the caucus quicker, easier and more secure for party chairs. However, the app was so riddled with problems that it caused major delays in the reporting of results, and an embarrassing setback for the party in its effort to modernise and improve the efficiency of civil elections.
The caucus debacle demonstrated both the perils of mismanagement and of efforts to digitise democracy without proper testing and investment, despite the positive intentions of the party. Here I’ll explain what a caucus is for international readers, why an app was used, why it failed to deliver a result and what lessons we can take from the experience of the Democratic Party in Iowa.
What is a caucus?
The American election is often described as a marathon, not a sprint. The New York Times estimates that 14 British or Australian elections or 41 French elections could have been held during the time of one US presidential campaign. The first step for voters is choosing a presidential nominee, through either a caucus or a primary. The majority of states in the US use primaries to cast their vote but some, such as Iowa, use caucuses.
In a caucus, voting for a candidate is a social affair: people congregate in halls, churches, high schools and literally stand up or raise their hands for the candidates they support. Caucuses are only open to party members and in a Democratic Party caucus there are multiple rounds of voting. A candidate needs 15% of the vote to progress. If, in round one, a candidate does not meet this threshold the vote will be held again, and supporters of other candidates can convince others to change their vote.
Why use an app?
Usually in a caucus, each local party chair will count the number of votes and report them to the Democratic Party. The party will collate the results and announce a winner. However, in the 2020 Iowa caucus, the Democratic Party opted to use an app, which was meant to enable the chairs of each congregation to enter their specific vote count into the app directly.
The idea of digitising voting in primaries and caucuses is not new. In 2016, an app built by Microsoft was successfully used in both Republican and Democratic caucuses in Iowa. Precinct chairs in both parties managed to report, validate and publish 95% of their precinct results within four hours. In March 2000, the Arizona Democratic Party ran its primary completely online: each voter had an individual PIN and once verified, could vote for their preferred candidate.
What went wrong?
The app used in Iowa’s caucus this year was built in just two months by a relatively unknown company called Shadow Inc., a Washington, D.C. based start-up with political connections. The development of the app cost the Iowa Democratic Party approximately US$60,000, according to state campaign finance records. The low level of financial investment in the project and insufficient time for development arguably paid off as should have been expected.
Instead of being made available via an app store, the app was distributed via a platform commonly used during the beta-testing stage for new apps. This by-passed the rigorous testing and troubleshooting processes involved in listing an app in the App Store or a similar platform. Party officials knew there were problems with the app before voting began, with some reporting that they were unable to even download the app before the vote due to multiple security measures that made the process more complicated. NBC News obtained messages from a group of party chairs on the morning of the vote. One chair said, “I gave up on the app,” while another said, “Just don’t submit your results!”
Officials were not provided adequate, or in some cases, any, training on how to use the app. During the vote, when party chairs attempted to enter their vote count, they were presented with an error message. As more and more party chairs reported the same problem, they defaulted to ringing or texting the Democratic National Committee to report their results. Some were advised to take photos of their results and send them in that way. Out of 1,765 precincts, only 624 people logged into the live version of the app and 439 (or about a quarter of precinct chairs) successfully submitted results using the technology, according to Iowa Democratic officials.
What lessons can we learn from this?
The app produced for this year’s Iowa caucus was a prime opportunity for the party to capitalise on earlier, sporadic implementation of voter reporting technology to prove the concept and illustrate that digital voting had the potential to be secure, streamlined and inclusive. Instead, management failures and a lack of investment in proper trials and testing meant that the app failed to perform on the day, leading to uncertain confidence in the result of the caucus — a critical error in a time of decreasing global confidence in electoral processes.
For some party chairs, this app was their first experience of digital democracy. The failure of the process led to the resignation of state’s Democratic Party chairman, Troy Price, and broad criticism of the party from across the academic field.
“The consensus of all experts who have been thinking about this is unequivocal,” a professor of computer science and law at Georgetown University told the New York Times: “Internet and mobile voting should not be used at this time in civil elections.”
The lesson policymakers across the world can draw from this experience is that for the successful adoption of new digital tools, they must not only be well-designed, well thought-out and properly troubleshooted, but must also be accessible to their users: whether they are younger and tech-savvy, or less versed in digital tools and mobile apps. Extensive testing on a national scale must be conducted prior to introducing new digital tools to democratic processes, given the risk of failure and the impact this may have on citizens’ trust in the process. The Democratic Party learnt a hard lesson: ironically, it is better to be conservative, rather than over-eager, when it comes to introducing new technology to a process as complex as voting.