In the field of design, one of the most useful paradigms is the activity. It might also be called the unit of behavior. It is divisible into subordinate activities, to some extent, and may be composed as a member of a superior activity.

A brief example will serve.

Cooking, as Always

If you are making a cake, it is unlikely you are planning merely to eat cake. You are likely causing the cake with the expectation that you (or whomever the cake belongs to) will enjoy it following some greater meal. You may be responsible for preparing or procuring the greater meal, or it may merely be an expectation.

The cake is a piece of a larger activity, either a meal, a party, what-have-you. Likewise, the cake making has smaller activities inside it. The icing. The layers. Icing the cake. Assembling the layers. Buying or bartering for the ingredients. Planning the cake.

Activities are Goal-oriented

Humans are goal-oriented, and activities are their attempts to achieve goals. Computers help humans complete activities. In some cases, computers can eliminate whole activities.

For any activity that a computer can eliminate, it ought be eliminated unless the human gains some value in completing it. That value could be the computer working on some less-desirable activity, learning from the activity, doing a better job than the computer, or something else.

These caveats are all goal-oriented. If you exercise to stay fit, having the computer do your exercise would preclude the goal, so eliminating that activity wouldn’t be sensible.

The goal is what’s notable. It’s what’s important. And now to the goal of this post, to talk about voting as an activity.

The Background to Voting

We’ve established the goal-nature of activities. What is the goal-nature of voting? Let’s step back and say what voting is. And as per usual, we can start with a dictionary definition and etymology.

From Wiktionary: vote: noun:

A formalized choice on matters of administration or other democratic activities.

From Online Etymology Dictionary: vote:

mid-15c., from L. votum “a vow, wish, promise, dedication,” […]

So to vote is to make a definite proclamation of one’s wish for some group choice. The goal being to establish the collective wisdom on the matter at hand. The best right ??? choice. Here we run into a problem.

We can’t really say the choice is anything but that which had consensus. Which is tautological. If the choice is unanimous, maybe we can go slightly further, but even then we cannot state it’s the best or right choice.

Now up the chain, to the agreement to choose based on voting. Again, we cannot say this is the best method of choosing, but merely the agreed upon method. If we had a better method, we would need not rely upon voting.

If, for example, we were choosing which key to use to open a lock, we would surely not vote. We would pick the key that fits the lock.

We are often faced with more complex choices than those with one-to-one correspondences. But, if we choose to join together in the first place, as a self-governing institution, for example, we do so under an agreement to have the institution run based on the consensus vote.

We may also provide for prerequisites, limits to the authority, in entering into this agreement.

The Integrity of Voting

Once we establish the vote, what is the nature of the activity itself? It is to have the eligible voters, in whatever allotment of votes they may have, cast their ballots. The ballots, tallied, give us the result. We have our choice.

We can imagine numerous ways to allot votes, to cast votes, to tally votes. We can even create perverse eligibility requirements prior to the agreement, in the form of prerequisites.

But once we have established the agreement, barring interim amendments conforming to the agreement, the activity of voting is fixed.

What I’m getting at is the fact that there is nothing inherent in the system of voting that gives it even a modicum of integrity.

If we are to have a voting system with integrity, it us up to us to provide it for a given institution.

The Voting Activity

Which finally arrives at the activity of voting. We have a fairly robust system, with a good bit of built-in integrity. This integrity arrives in the form of eligibility requirements for office, such as residency. In the form of allotment of votes, driven by a census. In the form of eligibility of voters, in terms of age, without restriction for the poor, and only removing the vote from duly convicted felons (we lose integrity here) and the mentally incompetent (we may lose some here as well).

But we lose a lot of integrity when we issue requirements that make the act of voting more onerous without concomitant gains. This includes extreme restrictions of the time and place of voting. It includes making the voting process needlessly complex, or requiring extra, unnecessary bureaucratic interactions.

If we approach the voting activity as an activity, to be designed for smooth and orderly action, we would not choose these severe limits on the times, places, and extra requirements for voting.

For example, take the issue of these draconian voter identification requirements being passed around the country. If they were based off of an existing, robust system for authentication, it would be acceptable. But it’s almost solely based off of the driver licensing scheme, which is already ineffective for a large number of authentication purposes.

We have a dire need for a good authenticating system for a variety of other purposes, but these too are neglected. As a society we lose non-trivial amounts of capital because of our lack of authentication.

In a similar manner, we have existing infrastructure (and could build more) that could be used for voting purposes, including ATMs and (where they still exist or where new ones may be added) phone booths.

Conclusion

Design problems are ubiquitous. The tendency to highlight design problems in software (both backend and user experience/user interface) is caused merely by the proximity we have to these systems and the immediacy of the relationship between the problems and their results.

If we design our larger systems with more rigor, we can do far better than we have. But we run into the same problems in software that we do in systems such as voting and banking. These include the same competing interests between marketing and engineering, for example.

It remains to be seen exactly how we can overcome these issues on the large scale. Continuing to examine these issues should prove fruitful, though.