Browsing the
individually scanned ballots
from Humboldt County’s Nov. 3 election, as
suggested
by my esteemed colleague, can be tedious and, frankly, a bit disheartening. Turnout was
reportedly
less than 22 percent, which makes these head-scratching fuck-ups all the more painful:
Were the instructions unclear? (There’s a shit-ton like the above, BTW.)
Okay, seriously, WTF?
There are no words:
This article appears in Reinventing Scotia.
Rock the Vote!
WOW. Are these the same people who didn’t follow the rule in elementary school not to eat glue and sniff markers?
Thanks for looking, Ryan.
It gets worse. Take a look at how many two sided ballots have only one side filled out, despite "vote both sides" printed on top and bottom. Take a look at how many "vote for three" have only one vote.
If anyone is interested, my draft report on the counting process is available at this location:
http://docs.google.com/View?id=dgxf37nt_155rw4spndw
Thanks for doin’ it, Mitch. It’s invaluable data, and I hope it starts a trend.
And you’re so right about the partial voting. This ballot is just one of many such examples. Still, better than not showing up at all.
A partially completed ballot doesn’t strike me as a mistake. In the last example, the person only voted in the harbor race — the most prominent and well publicized race. The person probably didn’t follow the school board races. I favor that person’s not voting over his voting based on how much he likes a person’s name, occupation or sequential position in the list of candidates.
Yes, AJ, you’re right. Many partially completed ballots just mean that a voter only wanted to vote certain races. But when you see one side filled out and the other empty, it makes you wonder. And when you see multiple "vote for n" contests but the voter has filled in one vote for each and every contest, it makes you wonder.
I doubt there’s any way to know for certain what portion of the voters missed or misunderstood the ballot instructions. But people involved in elections could certainly do design testing of different ballot approaches to see which resulted in the lowest error rates. For example, I wouldn’t be surprised if it turned out that showing examples of how NOT to mark the box actually increased the error rate. And I’d put the words "VOTE FOR UP TO THREE" in a full-height box to the right of all the candidates in a vote for three contest.