May 05 2014

Some notes on Early Voting numbers

Posted by Steve Ross in elections, Shelby County

Vote Where?

Vote Early?

Early voting is over and the numbers are in. The newspapers (here, here, and here) reporting that Early Voting turnout is a dismal 15% less than in 2010. The question they never even approach is why?

Days before the first day of Early Voting at Satellite locations, Democratic Candidates Henri Brooks and Deidre Malone charged the Election Commission with voter suppression for reducing the number of days satellite locations were open as compared to the May 2010 County Primary.

The Election Commission says they reduced the number to reduce the cost of the election…and that the additional days of early voting at Satellite locations could be absorbed by the downtown location.

One of the issues brought up at the presser with Brooks and Malone was that the downtown location wasn’t centralized enough or accessible to adequately cover the difference in the number of days Satellite locations were available…something the Election Commission seemed willing to look into for future elections.

So there are really two questions here:
1. Did the downtown location absorb a significant amount of the early voting traffic?
2. Did the absence of Early Voting days at Satellite locations suppress the vote?

To find that out, we have to get into the numbers.

May 2010 vs. May 2014

Early voting at Satellite locations in the May 2010 election began on Monday, April 19th. Satellite location this year opened on Friday, April 25th. That means there were four fewer days for satellite locations this year than in 2010.

This year, there were 4707 fewer people that participated in early voting. Here’s the breakdown by location types.

2010

2014

Diff +/-

Downtown
1898
1991
+93
Satellite
27982
23300
-4682
Absentee
387
263
-124
Assisted Living
738
744

+6

Total
31005
26298
-4707
Average Voters per day
2798
3883

 

First, let’s answer our two questions:
1. Did the downtown location absorb a significant amount of the early voting traffic? Not Even Close

The downtown location underperformed versus 2010 big time because for four MORE days it was the only game in town. For those four more days it brought in less than 25 more people per day. That’s hardly absorbing anything of consequence.

2. Did the absence of Early Voting days at Satellite locations suppress the vote?

Before I go there, let’s take a look at some data.

Looking at the table below, it seems pretty clear that the final days of early voting in 2014 were more brisk than in 2010. This is most likely due to the fewer number of days.

But by the end of Early Voting, an 8100 vote gap was narrowed to just under 4700 votes.

Cumulative Satellite votes per day. 2010 vs. 2014

 
1
2
3
4
5
6
7
8
9
10
2010
2209
4301
6184
8091
9979
10830
13390
16794
21012
27982
2014
0
0
0
0
2434
3896
6663
10735
15858
23300
Diff +/-
2209
4301
6184
8091
7545
6934
6727
6059
5154
4682

Early Voting Statistics Click to embiggen

Early Voting Statistics
Click to embiggen

Its interesting to me that the difference between the number of people that voted early in 2010 and 2014 is almost exactly the same number as the difference between satellite location participation in those two elections. At first glance, it would seem the vote was suppressed by that number of votes.

The fact is, with 40% fewer voting hours at satellite locations, satellite voting was only 16.7% less. On an hourly basis, there were actually about 70 MORE voters per hour in 2014 than in 2010.

Had the full 81 hours been given at that pace the final satellite tally could have been about 5700 MORE voters in early voting, which means we would have outpaced 2010 by about 1000 voters.

So its a bit disingenuous to say Early Voting is down 15% without noting the number of hours satellite locations were open is also down 40%.

All told, early voting performance for any May election was pretty good. Was it good enough to warrant 10 days of satellite locations? Probably not.

The August 2010 election brought over 90,000 early voters. That election absolutely needed to have 10 days of satellite locations.

Suppressive or not?

Back to our second question. Was the decision to limit satellite locations to 6 days inherently suppressive? It certainly didn’t help early voting turnout, but lets not pretend that people were flocking to the polls on any day other than the last day of Early Voting.

That said, if Tuesday night rolls around and the vote total is close to or greater than 2010 (they’re projecting about a 10% decline from 2010), it’ll be pretty clear that the Election Commission’s projections on this were efficient from a financial standpoint, but off the mark from a “serving the public” standpoint.

Six days is probably too few, and ten days is probably too many.

There are lots of things the Election Commission can look at, including underperforming early voting locations to get the right mix of May early voting locations. For instance, one location only saw 333 votes cast in 6 days which is low by any measure.

But since this election really only happens once every four years, its a crap shoot. Hopefully for August they’ll take the savings from this election, and make early voting more accessible to voters.

Of course, you’d never know what they’re planning since the Election Commissions Minutes and Agendas page is 4 to 6 months behind. (sheesh, seriously people?)

I’ll take a deeper look into the final election results on Wednesday and Thursday.

Stay tuned.