Wednesday, April 18, 2012

War on Polling Places

It may not be as dramatic sounding as the media's phrase, "War on Christmas," or many of the other wars on societal issues, but as we prepare for more elections, we're reminded of the constant war on polling places.

Selecting polling places is a no-win endeavor.

For instance, in April 2005, the election featured a question on same-sex marriage.  I received several complaints from voters that some of our polling places were churches, potentially influencing the outcome of this vote.

Then, in September 2005, we had a special election for a sales tax that was directed to schools.  I received a similar number of complaints from voters that some of our polling places were schools, potentially influencing the outcome of this vote.

We used the same polling places for both elections.

Most of our polling places are donated space.  That's important because one thing I hear often from our county manager is how expensive elections are.

They are expensive.  But that expense is relevant if you are comparing the cost to zero.  Merely having an election is expensive because it's an event for, in our case, 360,000 people.

Consider that voting machines are a sunk cost, polling places are nearly free, election workers make minimum wage (if that) and that our most expensive part of the election is the bundle that comes as an advance ballot by mail (ballot printing, envelope printing, and postage).

Of the 284 Polling Places used
in the 2008 Presidential election,
we spent $5,000 in rent because
the majority of them were free.
Of course, fewer polling places with more election workers squeezes efficiencies.  Instead of having 500 polling places with 3 workers, we could have 300 with 5 workers--same number of workers, but less expense in transporting voting supplies, fewer overall supplies, and fewer machines needed.

We do that already.  Ideally, we'd match polling places to precincts and we have about 550 precincts.

Instead, we utilized 284 in 2008, down from 286 in 2004.  Incidentally, that's not two we lost--we lost 100 and found 98 new sites, for a net reduction of 2.

This year, we'll probably be at around 275.  We'd like to secure the same location for voters for a string of elections so they aren't moved frequently (another complaint).

Johnson County is fortunate that we have many newer churches and facilities that can be used as polling places, that are compliant with the Americans with Disabilities Act, and have adequate parking.

Parking is vital.  Putting 2,000 voters into a polling place with 40 parking spaces causes problems (another complaint).

So, what if we went nutzoid, didn't "pin" a person to a particular polling place and had, essentially, the advance voting model on election day?  In other parts of the country, this has been tried and is often termed "vote centers."

We have a pretty good feel for the number of persons who would vote in a presidential election and how many would vote in advance, so, roughly, we would need to take care of about 200,000 voters on election day.

If we had 60 locations, that would average about 3,300 persons voting at each site.  That's more than we've had at any of our advance voting locations on any day, but it gives us a feel for the size of the facilities we would need.  Our advance voting locations have handled 2,500 voters before with about a 45-minute wait.

So, that's leap one, that 60 would be okay and that 45-minutes, on average, would be an acceptable wait to vote.

About those 60--they'll have to be networked together so we make sure someone doesn't vote at one place and bop on to another.    First, though, where exactly are these 60 places?

We might be able to secure some very large churches and community centers, but we'd probably have to rent facilities:  likely, hotels.  So, when we look at the network costs, although there may be some wireless options at times that are cheaper, we'll often have to go through the facility itself.

We have one person negotiate space with our polling places and our advance voting locations.  The polling places are relatively easy but when we get to the advance sites--with rent, contracts, insurance, etc.--it becomes extremely time consuming.  We have no resource, or support, to identify and negotiate with 100 potential locations to get us to 60 that work.

Assuming we could get through that knothole, though, this idea looks pretty good.  We'd still have about the same number of election workers, but we would gain efficiencies in machine use.

Our rent costs, though, would increase from about $5,000 for an election to a couple hundred thousand dollars.  Again, elections are expensive when compared to a cost of zero, but compared to real-world costs, elections are a bargain.  We'd hit the real world if we sign leases with hotels and other for-profit locations.

Results would take longer because it would take about an hour to close all of the machines (they'd have every race in the county loaded instead of just those at one polling place).  Of course, the line at 7 p.m. would be longer, too, so the machines may not be shut down until about 7:45 p.m.,  with cards leaving the facility at about 8:45, arriving at our location for uploading perhaps by 9, with uploading done by midnight.

We get skewered when our results for a presidential election aren't up by 10:30 now, even though we're traditionally the first in the area and the first large county in the state with final results.

Hmmm...this reminds me of a voting task force meeting I was in a couple of years ago when a professor from Washburn provoked, "If you thought about it, starting from scratch, would it make sense to design a voting system where you had hundreds of polling places where people could only vote where they were assigned?"

"Of course not!," I thought.  We were all nodding our heads in agreement with the wisdom this question brought.

But, as I sat and thought more about that question, and in pondering some of the details I just typed, I think my answer would be yes.  I almost think that successfully moving from that model would require something even more dramatic than vote centers and that would be a single voting location for the county on election day.

Of course, there are some who might say that a single location would be possible if that location was a website and we had Internet voting, but that's a whole 'nother discussion and blog post.

A single location is equal part crazy talk and genius, but it probably leans eventually to the crazy talk side.

For now, having opened up a topic without a clear solution, we're off to assigning voters for this fall to polling places.