Greg's Opinion Greg's big blog of whatnot

18Feb/141

Early Voting – 2014 Primary Edition

It begins again. Now is the time for Early Voting for the March primaries. I got my ritual out of the way first thing this morning. Locations are below. Hours are below that.

February 18 - February 21: 8:00 a.m. - 4:30 p.m.
February 22: 7:00 a.m. - 7:00 p.m.
February 23: 1:00 p.m. - 6:00 p.m.
February 24 - February 28: 7:00 a.m. - 7:00 p.m

As a followup of sorts to the Substantially Similar name affidavit posts, I recently got my hands on the last portion of my request from the County Clerk's office: the match file to show how the voter file compared to the DPS database. Media reports prior to the 2013 election had Stan Stanart saying that 40% of the names on the voter file did not match. I databased the file I received and did some crunching to see what turned up and I got 20.9% of the records that were not a perfect match. Of course, I also had a voter file of 1,871,369 voters compared to the reported 1,967,881 reported on the county's canvass of the election. I've put in my questions to the folks at the Clerk's office to see what might explain the differences and haven't heard back yet. It may very well be that I have a file that was run after the election, or there could be other things to account for the 100,000 or so names that didn't match. If I were to assume that the 100k missing records would ultimately be mismatches, then the countywide share of mismatches would be about 25%. For all I know, I may be comparing apples-to-oranges. But since it's the only information I've got from the County Clerk, I'm going with it for now.

Now, following my earlier logic, I would assume that an odd-year election with high-propensity voter turnout would lead to less than whatever the DPS match is. As a hypothesis, I might expect the share of voters who should have been required to initial a similar name affidavit to be somewhere in the 10-15% range for 2013. I'm still in mid-crunch for the overall numbers. But early indications are that the actual share who should have initialed an affidavit are right around 20% - essentially the same as my database comparison.

A few things are turning up interesting from this work: for starters, I can gauge which locations did a reasonably decent job of efficiently capturing initialed affidavits. If the "should have" calculation is, say, 100 voters for a given day ... and the "actual" count of affidavits is 90, it's reasonable to assume a 90% efficiency for that location.

Complicating things are locations like 128P - Pasadena's Harris County Court Annex. This location should have had 10.7% of their voters initial affidavits. In actuality, they led the entire county with a league-leading 24.5%. In other words, they were asking voters to initial affidavits for no justifiable reason. It may be completely understandable that each location would have a few voters be asked to initial an affidavit mistakenly. But I would also expect those to be more than offset by voters who should have initialed, but did not do so. Under no circumstances that I can comprehend, would you have a situation where almost three times the number of voters who needed to sign, did so. As a matter of full disclosure, the location I worked at had 4 days where our "actual" exceeded the "shoulda" count and we had one day that tied perfectly. On the whole, Bayland Park clocked in at 81% efficiency for all of Early Voting while Pasadena clocked in at 229%.

I will note that my earlier posts were passed around by the Clerk's office to Election Judges. And when I voted today, my "substantially similar" name comparison was properly identified and I initialed the affidavit. When we're done with the primaries, we'll have a few new things to look for: was there progress in the administration of the law ... and what do the name matches look like when broken down by party.

In the meantime, I should have all of my number-crunching done for the "shoulda" vs "actual" affidavits completed this week. Fun stuff.

Related Posts:

Comments (1) Trackbacks (0)
  1. Great work here, as always.


Leave a comment

Trackbacks are disabled.


Fatal error: Call to undefined function xwp_footerx() in /nfs/c01/h12/mnt/4748/domains/gregsopinion.com/html/wp/wp-content/themes/lightword/footer.php on line 18