Technology Provides Tools, Not Panaceas

I take issue with Keith Spencer's clear implication in his Salon article that technology has no role in the elections. Two of the three objections he cites to the "computationalist mindset" are actually objections against the use of computers: "First, it enables malfeasance and incompetence.... Pen and paper, on the other hand, are tried-and-true tabulation methods that are not hackable. Second, computers are opaque...A layperson cannot audit the code of an app or website and determine if it is being manipulated."

This is missing the point. Technology offers tools which in many cases can be put to good use, but which are just easier to employ in short-sighted or nefarious ways. It is often the human systems which design and use these tools that are not carefully planned and thus enable vulnerabilities, accidents, or malfeasance. The lack of care in establishing these human processes may indeed be partly due to the fetishization of technology as an automatic panacea requiring no further thought, when in fact any social process needs to be carefully planned to ensure it achieves its purpose. In the case of elections, even voting by pen and paper can have issues with mis-tallied votes and missing or destroyed ballots, not to mention disenfranchisement at or even before the voter arrives at the polls.

Technology has the potential to help address these issues. Two requirements seem key: first, that the technology itself be easily auditable, and second, that the voting process be transparent. To achieve these ends for a public service of such civic importance as voting, human participation needs to be an inherent part of the development and use of the service. For example, responsible design decisions might be first, that the technology be open-source, and second, that the vote records be publicly visible by a mechanism such as the blockchain. Open-source code allows a wide audit of the technology to identify vulnerabilities before they affect an actual election. While this audit (like any audit) can only be carried out by experts, by making the code open-source the experts are not confined to a particular company with possible conflicts of interest or incentives to cut corners. Using a technology such as blockchain allows any individual or group independent of the government (albeit with sufficient computational resources) to verify the vote tallies, thus eliminating doubts about the results. In this way, technology could give us more faith in our elections and could potentially even expand the franchise by allowing people to vote from more places, reducing or eliminating the long waits at polling stations which make voting inconvenient or impossible in some precincts.

These ideas are just examples of how technology can help humans conduct a process as complicated as elections while allowing us to remain in control of and be confident in these tools. Obviously, any e-voting solution needs to address numerous other design issues on both the human and technological sides of the system: how will voter eligibility be verified? how can voters track their vote while preserving anonymity? how can the open-source platform be protected from attacks?

Properly designed technology can certainly help us conduct elections, just as it can help many other aspects of our lives. The harder part, however, is establishing the human systems that use the technology so that they're fair, equitable, and serve the common good. That is where we fall short.


Popular posts from this blog

Sharing the Dream

Leaving on principle or staying for change?

Rebooting Critical Exponent!