DIY: Brian Newby on Printing Ballots In-House

[Image courtesy of talencio]

Brian Newby’s latest at ElectionDiary [“Printing Money (Savings)“] describes Johnson County’s move into an experimental new phase in election prep: printing their own ballots.

The step comes in response to Brian’s and his team’s realization that huge numbers of ballots were going wasted in each election:

Our thought process started as we evaluated, again, the need to take carts and carts of paper ballots to our satellite advance voting sites. We have to do this to have quantities of the nearly 1,500 ballot types in August in case there is a need to issue a provisional ballot.

We don’t even have 1,500 provisional ballots from our four in-person advance locations, let alone from one of them and, of course, we have no idea which ballots would be needed, so each satellite site has a fair quantity of each of these 1,500 types.

Then, most of these ballots go unused and, after election day, are destroyed. The ballot carts come back to our office and sit, taking up space, until the next election.

That realization prompted a decision to experiment with ballot on demand:

We started 2014 … with the crazy idea that we would not send the carts to two locations–the Northeast Johnson County office and the Great Mall of the Great Plains. Instead, we would equip ballot-on-demand printers and technology to print necessary ballots there.

Our theory, if this was a major fail, was that voters could drive the five miles or so to the next closest advance site, equipped with the carts, if the printer didn’t work. We knew the technology had been used reliably elsewhere, but power failures before because of August thunderstorms seemed like the biggest risk.

So far, more than so good! This plan allowed us to send more voting machines to the smaller Northeast facility, kicking out 10 carts and putting in 3 more machines.

Not surprisingly, Brian and his team used that success to attempt another ambitious goal:

Then, armed with the printer order, our overzealous staff decided to jump the shark and print all of our advance by mail ballots on demand …

We’ve had a few growing pains. Some of the pages stuck together in one batch, leading to 4 ballots being sent with no back printed, but those voters have new ballots and our printing team has new proofing procedures.

We used the ballot on demand printers for our public test process so we’d be sure we didn’t send out thousands of ballots that wouldn’t scan. We also used the technology as a trial in April and with our recent Roeland Park election. We printed all the paper ballots for Roeland Park in-house, although that was a very small number.

In the end, the printer project became much bigger than we first envisioned but as a result, there is an overall ball of stress that has traditionally joined us for each election and is noticeably absent.

Here’s the kicker – it looks like it’s saving the county money:

By my back of the ballot envelope figuring, we’ve avoided about $15,000 of unnecessary printing in this election. These two 2014 elections (August and November) should return at least half of the upfront expense, and we expect the life of the system to extend well beyond “break even.”

But when the dust settles, we’ll also compare labor costs to 2012’s ballot-labor costs. Intuitively, based on number of individuals in the back this year vs. then (and the ballot distribution has been almost identical), we likely have saved considerable temporary labor expense as well.

Even better, it’s easing (somewhat) the workload of ballot staff:

We’ve implemented a few new processes that have, in different ways, led to a few extra hours of free time for our ballot team.

Plus, those who balance our advance voting totals have been been going home by 9 each night rather than midnight as in years past. That ends up resulting in weeks of 14-hour days rather than 16-hour days, and that difference is huge.

[No word yet on whether the new procedure will result in the consumption of fewer oatmeal creme pies at Brian’s office or just up the rate per hour.]

As I’ve documented several times over the years, the process of designing and printing ballots by outside vendors has not been without problems; while this solution creates new potential problems it does have the promise of putting a key aspect of election prep – the ballot itself – more squarely in the election official’s control.

Thanks, as always, to Brian and the Johnson County team for their continued leadership on issues like this … even better, I know Brian will post updates on how well (or not) the experiment works in the weeks and months to come!

Stay tuned …