Thursday, December 13, 2012


Apparently, Christian Slater, the actor, had difficulty voting in the presidential election.

I first heard this during a panel session at PEW's Voting in America 2012, an invitation-only conference of election administrators, elected officials, and smart people that I attended earlier this week in Washington, D.C.

The comment came during a discussion by Robin Carnahan, retiring Missouri Secretary of State.  She used the actor's example to point out registration issues in elections.  An inconsistency with his registration, as the story went, kept his vote from being counted in Florida.

Registration process changes are but one of many things considered as answers to the infamous President Obama quote on election night related to, "We've got to fix that."

(Random thought #21--when boarding my flight out Sunday morning, I had a pleasant stop and chat with Senator Jerry Moran.  Obviously, one discussion with just one Senator is a small sample size, but it was evident to me during that conversation that the "fix that" comment was not anywhere near top of mind in the U.S. Senate).

(Random thought #186--you may wonder why random thoughts are numbered, if they are random, and I must point out that they are numbered solely for the convenience of the reader).

Anyway, the session before the Christian Slater example was much more impactful, in my opinion.  Heather Gerken led a panel to discuss the creation of PEW's Election Performance Index that will be rolling out soon.

Heather started this all with her book, The Democracy Index, and the whole concept here is using data to measure performance.  It gets to the foundation of Stephen Covey's concept of, "Put something on a wall and measure it, and it will improve."

Heather is a rock star when it comes to elections--not to be confused with another Heather, Heather Smith of Rock the Vote.  She spoke on Monday and is very impressive as well.

They are two of my favorite Heathers.  A third, not at the conference, is Heather Taylor of E-Consultancy, who posted about social media and elections coincidentally the first day of the conference.  I highly recommend following her on Twitter.

There, three Heathers and Christian Slater (sort of) together over two days, but this wasn't a movie with a bomb, although it was broadcast on CSpan3 and we were beginning to evacuate the Newseum just before the conference began because of some safety issue that was quickly resolved.

I'll speak to one of the other panels, related to social media, in a future post, but sticking to the Performance Index, it is something that we want to nudge along in Johnson County.  I'll have some discussion about the Index itself later as well, but Heather Gerken's book is a good read in the interim.
Panel Discussion on the PEW
Election Performance Index--Former
(and outstanding) EAC Commissioner
Ray Martinez, Michigan Election
Director Chris Thomas, Smartguy
MIT professor Charles Stewart, and
PEW Index Champion Zachary
Markovits.  Behind the podium
is Heather Gerkin--you see her boots
only, presumably used to kick
tails and take names (and
The great thing about the Index is that it is happening.  The bad thing is that no matter how it's prefaced as a discussion-starter or preliminary, first impressions are lasting impressions.  When all of us see the data, we immediately look to to see how our state has done. 

The next debate is whether these are the right measurements or are they equally weighted or what the most important measurement is or even if the measurements themselves are accurate or if we are measuring apples-to-apples.

"Wait time," for instance, is something measured but it doesn't break down if the voter voted in advance or at the polls.  And, it's self-reported by the voter, so it may actually be the perception of the wait, not the actual time.  That perception may be more important, or less, but this gives you a tiny taste of the discussions that might spin from the data.

A trendy job these days is that of Data Scientist, tracking "Big Data," and breaking it down into insightful bites.  We need a Data Scientist at the Johnson County Election Office.

For that matter, we need a financial and strategic planning manager, an outreach manager, and at least three more employees to simply assist with election blocking and tackling so we aren't dependent solely on one person without a backup at critical points in the election cycle.

We haven't been able to get those basic needs filled, having the same number of employees for the last 20 years despite seismic changes in elections administration, although we're going to take another swing that that during the next budget cycle.  Still, a push for a Data Scientist would only result in eye rolls--we'd have a better chance getting another trendy (and not needed) position, a Sustainability Manager.  We need the real positions to ensure we can sustain administering elections, so I likely won't push for a Data Scientist.

To me, that's the biggest risk with the Performance Index.  The benefits, I believe anyway, can't be disputed.  But without devoting a resource to really get our fingernails dirty with the data, we run the risk of the wrong people seeing the wrong metric and drawing false conclusions, good or bad.

It's a good start, but I'm sure I'm about 60 days or so from getting a call from our local radio news station wanting a quote on how Kansas fared on the Performance Index.  In some ways, the Index truly will be the bomb I referenced earlier and if my anecdotal "Fix That," research is any indication, we'll be talking about the Index much longer than a line from a presidential acceptance speech.


Nicholas said...


Thanks for the blog. I like the direction you are headed. I am bewildered that more elections offices do not have a "data scientist". I used to work for the Leon County, Florida Supervisor of Elections Office as the Demographics/GIS Manager. They are one of the few elections offices in Florida that has a dedicated person on staff who looks at this stuff. I am no longer there, but continue to work on election administration questions as a PhD student in Public Administration. I live in Miami now and I do not believe they have anyone on staff making predictions or doing performance metrics using all of that great data. I will contend that this cannot be someone from your traditional IT staff. It has to be someone who has a mix of social science and technical skills. It is as much an art as it is anything else.

I am an election geek as well. I hope to keep in the conversation and helping to move elections forward as we technologize. I plan to solve your Magic Math Machine problem. I will keep you posted.

Nicholas Martinez

Election Diary said...

Nicholas, thank you! If you solve the Magic Math Machine problem, I bet, if she were around today, Sister Shirley would give you a happy face sticker!