Such was the case Tuesday, where Roeland Park picked a new city council member after a resignation.
Roeland Park is the only Johnson County jurisdiction that doesn't appoint replacements on its council. This was the second special election in Roeland Park in 2014.
Like the previous special elections, it was a one-ward, one polling place election.
The election also wrapped by my first 10 years of elections as election commissioner. We've had 60 elections, but only 20 of them were planned.
Averaging 6 a year, we already have 5 special elections planned for January 27--five different school districts will be having mail-ballot elections on the same day.
This election came during a busy week, as I also needed to prepare for a presentation to a Kansas House and Senate Committee on Friday. This committee is looking at data related to the potential of moving spring elections to the fall, either in even or odd years.
I've been advocating moving the elections to the fall of odd years. I've discussed the reasons before, but, summed up, I believe turnout may increase and I also think it could be done in conjunction with requiring schools to have a student-holiday on election days to allow consistent use of schools as polling places.
So, the preparation of that presentation was on my mind on Tuesday as we waited to see the turnout--more than 20 percent, as it, well, turned out.
Hmmm. That's a higher turnout than spring elections (that include Roeland Park).
We got to thinking. The turnouts for special elections usually are higher than we expect, at least at the polls. Mail-ballot election turnout has been dropping over time, but the turnout still is very good.
We put it all on a slide, below, and note with the red arrows the lowest turnout of any special elections. The green arrow shows the highest turnout of the last four April elections.
The lowest turnout of a special election is still higher than the best turnout in April.
We've read criticisms of the idea of moving spring elections that there is no data to suggest turnout might increase if elections were moved from April to November.
That's true. But this data supports that elections held any time BUT April have a higher turnout.
It's fascinating, really, because the quick follow-up question is why?
It could be that these elections stand on their own, and get individual attention, even individual outreach attention on the mail-ballot side. The why is worth exploring, but the data is fairly conclusive that April turnouts are the low achievers in the crowd.