[Image (c) 2006 Sidney Harris]
In less than a week since Election Day, millions and millions of words have been written/spoken/typed about what's necessary to address problem of long lines - probably as many as many words as voters who cast their ballots in the 2012 election.
We have already seen just about every possible solution suggested in that time - all of which boil down with either changing one or more elements of the current system or increasing our commitment to it.
But regardless of the approach, no election reform will succeed in attacking the problem of long lines unless it is cognizant of the need to understand and respect basic math.
Indeed, at its root the effort to avoid long lines is a straightforward math problem:
(Number of voters) x (Time per voter) ≤ (Polling hours)
To put it another way, the overall goal is simple: make sure that the supply of voting is greater than or equal to demand.
Of course, this simple inequality masks far more challenging calculations. Take number of voters, for example. Voters do not approach a polling place equally spaced on a conveyor belt; rather, they arrive at different times of the day and tend to bunch in "rush hours" in the morning and evening. Moreover, we rarely know how many of them there are going to be - back in July, I talked about how a mistake in anticipating turnout in Anchorage was disastrous in a city election - and thus having a good idea what demand will look like on Election Day is vital.
Nor is the time per voter a simple calculation. Every voter comes to the process with different skills and abilities, meaning that "time to vote" can be affected by reading skill (and language), physical and cognitive ability - even the level of preparation in advance. In addition, it is a function of the actual act of voting - how easy it is (or isn't) to read and make choices, how long the ballot is, and the physical environment for voting. All of these factors combine to produce a "time to vote" that is almost as unique as the voter herself. [This is also where the other considerations of voting - equity, transparency, security etc. - become part of the process as they affect time to vote. Note also that I'm not addressing convenience - the "problem" of voters having to wait in line during polling hours - but rather the bigger problem of supply exceeding demand.]
Polling hours are simpler to calculate, but still have their own wrinkles. For in-person voting (either on or before Election Day) it's basically as simple as tallying up the number of hours the polling place is open; for absentee and vote-by-mail, however, it necessarily includes an understanding of how long it takes for ballot to reach a voter and be returned, whether by mail or in-person at a dropbox or other ballot collection center.
Note that problems with lines usually emerge when one or more of these factors is out of whack. As Whitney Quesenberry observed in the comments of last week's post-Election Day post, higher-population urban and suburban areas tend to be at greater risk of lines, given their increased voter volume; absent another problem, smaller communities usually don't have enough voters to clog a polling place. When time per voter begins to push the limit - because of a long ballot or a shortage of voting stations - lines can also result. If polling hours are too short, even a reasonable number of voters can find themselves in long lines. And in a place where all three factors come together like South Florida, where thousands of voters slogged through a 10-sided ballot after a reduced early voting period - you get the kind of epic lines referenced by the President in his speech.
Of course, like many mathematical formulae, the problem is easier to describe than the solution. What makes the science of lines so challenging is that all of this math can't be done in a factual vacuum; we don't know how long any of this takes unless we either physically observe voters as they vote or devise methods to capture data that can be analyzed after the fact. That's why I included the famous Sidney Harris cartoon above; in many ways, data on the voting process is the "miracle" we need to occur in order for our equation to balance.