Tuesday, September 30, 2014

Election Resource Allocation: A Real Issue, A Tricky Solution

As a lot of you recall from the last three presidential elections, election resource allocation is an issue in many states.





The charge has been that resources are allocated in a partisan manner, specifically that wealthy white neighborhoods get more workers/machines/etc. than poor black neighborhoods:
Three states that featured some of the longest wait times on Election Day 2012 are routinely failing to have as many voting machines and poll workers as they are required to -- and the shortfalls are most negatively impacting minority voters, according to a new study.
"We need to fix that," President Obama said as he claimed re-election victory. There's been no action, of course, in the hyper-partisan atmosphere of 21st Century DC. But eventually there may be, and as one who works in the trenches of elections I worry about unintended consequences.

Here's some stats that worry me:
In Maryland, just 11 percent of precincts studied by Brennan met the state's standard of having one voting machine for every 200 registered voters.

In South Carolina, a study of 16 counties showed just one-third of precincts in those counties met the state's standard of having three poll workers for every 500 registered voters. It also found 75 percent of precincts didn't meet the standard of having one voting machine for every 250 registered voters.
They worry me because, of course, they indicate insufficient resources. But they also worry me because they hint at a one size fits all solution which would work poorly here in my community.

Note that the standard used is registered voters. That assumes that registered voters translates directly into turnout, and does so in a flat, linear way across different types of elections. It also assumes that every registered voter is an equal workload.

Let's look at my own precinct, Iowa City 11. It's atypical, but that only helps illustrate the problem. We're about 80% students in downtown apartments, and 20% working class folks and some students in the Miller-Orchard neighborhood. (We used to be the Roosevelt School neighborhood till we lost the school.)

In the 2012 primary, we had 1409 registered voters. By the South Carolina standard, we would have had nine workers and six machines... for nineteen voters. Because our primary is in the SUMMER when school is out. A fact that might be relevant if you're writing a law.

Jump ahead to November. Registration has climbed to 2107 - at the beginning of Election Day. So the resources increase, by South Carolina standards again, to 15 workers and nine machines. There's not really ROOM for that in most polling places, but let's continue.

So registration has increased by half again between June and November... but turnout has jumped from 19 to 709. Young voter turnout is far more volatile and variable than any other demographic, and this is just an especially dramatic example.

Yet it's likely that legislation would be drafted to cover "federal elections," meaning both the primary and the general. And if the standard is based on "registered voters," the same level could mean workers are bored to tears in one election and overwhelmed in the next.

And some voters are simply more work than others. Of those 709 voters in my precinct in November 2012, at least 183 were election day registrations. An election day registration is far more work, for both the voter and the poll worker, than a voter who shows up registered and ready. We account for that when we assign workers... but would a one size fits all federal law let us?

The state laws cited above also, unless the fine print of the law says otherwise, fail to take into account the registered voters who have voted BEFORE Election Day.  Johnson County is the early vote champ of the state, with 58% of our total vote cast early in 2012. And we take that into account when we assign workers.

A federally mandated, registered voter based standard would leave us feather-bedded and waste loads of money, because the most expensive part of running an election is the workers, and we'd be hiring workers to wait on voters who voted a week or a month ago.

And in a final irony, even if the intent of a federal standard would be to equalize resources, it would end up skewing them in a different way. Absentee voting rates vary dramatically by state, by county, and even by precinct. Within our county in 2012, the rate ranged from 26% of registered voters voting early in small town/rural Oxford to 64% in Iowa City Precinct 2, dominated by the large Oaknoll senior living complex. So with a registered voter standard, Oxford would be under-served compared to Oaknoll.

There's so, so much else to think about. How long are each state's voting hours? Can you push the same number of people through with the same number of workers in two, three, even four hours less time? How long are the ballots?  (Just a few days into voting it's already obvious that people are taking MUCH longer in the booth with our long general election ballot than they did in the June primary with just a couple contested races.) How complicated is the check in process?


Sorry to not solve anything here. But any federal law is going to have to take all those things into account and more. And if the federal law throws it back to the states to set the standards, well, isn't that the problem in the first place?

Iowa law is relatively limited on this subject. We have a minimum three workers per precinct, with some level of party balance. Even that can be challenging, if your workforce has to be evenly balanced and your community isn't. Have I mentioned our office needs some Republican poll workers? (No idea how Sioux County finds Democrats.)

What guidance there is, though, directs auditors to base things like ballot printing and delivery on past similar elections. So if your primary is in June when a huge chunk of your community is gone, you can base it on past turnout rather than on registered voters. (Voter list maintenance - that's a whole `nother post...)

Even that can be challenging - in a downtown Iowa City precinct, how similar is the 2014 Senate election to the 2010 21 Bar and oh yeah governor and some other unimportant stuff election? But at least it allows us to take local realities into account. Maybe some type of federal legislation can address the very real, and yes very partisan, issues while still leaving some leeway for the locals.

No comments: