Trustworthy Democratic Technology

Over the last several decades, information technology has transformed the electoral process, which is perhaps the most foundational process upon which democratic societies are built. Computers are gradually replacing the manual parts of the democratic process. For example, rather than hand-counting a vote for days, Microsoft Excel is used to tally results. Or, more locally relevant, a UNIX program is used to compute the seat assignments to parties in parliament, as has been done since 1962 here in Denmark.

Some of these changes make the overall process more efficient or economical. However, there is a risk of the process becoming less trustworthy. The deployed technology tends to be complex, and is therefore prone to programming error and vulnerable to malicious attacks. This, in turn, has an adverse effect on the very foundations of democracy. Voters are less likely to trust the electoral process, which inevitably leads to lower voter participation and cynicism. Consequently, virtually all e-voting researchers, hackers, and activists are against the introduction of computers in the democratic process.

The hypothesis of this research project is the following: Is it possible to modernize the electoral process, while balancing the trust of the people on the trustworthiness of the deployed technology?

In order to provide evidence in support of (or to refute) this hypothesis, we propose to design a rigorous software engineering principle, which we call trust by design, that reproduces the trust�0;1F;instilling elements of the conventional democratic process in the new electronic infrastructure of elections. Together with our industry partners Aion and Siemens, we will develop electronic election technology based on the trust by design principle. Finally, working with Århus, Copenhagen, and Frederiksberg Municipalities, we will empirically evaluate the technology in order to analyze the social, political, and cultural implications inherent in the digital transformation of the democratic process.

The result of this project will be a computational democratic process and a reference technology platform for electronic elections that e-voting researchers, hackers, and activists the world-over will either (a) acknowledge as being trustworthy, correct, and secure enough for certain kinds of elections, or (b) we will refute the hypothesis and permanently close the door on the use of computers in the democratic process.

Regardless of the outcome, this research will provide decision makers, in Denmark and elsewhere, important and invaluable insights for how to modernize the democratic process without jeopardizing our fundamental democratic principles.