[EL] Republicans Allege Obama Admin. Not Doing Enough to Help Military Voters

Thu Oct 4 19:35:21 PDT 2012

 Michael -- What counties are missing?  If Wake, Forsythe, Greensboro and Charlotte (don't remember the county names for the last two) are part of your unknowns then it could foul up your numbers, if you remember the work you and I did on the EAC survey.

Kimball Brace 
Election Data Services, Inc. 
6171 Emerywood Ct 
Manassas, VA 20112-3078 
(202) 789-2004 or (703) 580-7267 
Fax: 703-580-6258 
Cell: 202-607-5857 
KBrace at aol.com or KBrace at electiondataservices.com 
www.electiondataservices.com 

NOTE: WE'VE MOVED: Please update your records. 
==============================================

-----Original Message-----
From: Michael McDonald <mmcdon at gmu.edu>
To: law-election <law-election at uci.edu>
Sent: Thu, Oct 4, 2012 9:46 pm
Subject: Re: [EL] Republicans Allege Obama Admin. Not Doing Enough to Help Military Voters

I've looked into these claims for North Carolina, and others can verify my
analysis with the information publicly posted by the state here:

ftp://www.app.sboe.state.nc.us/enrs/ 

The Military Voter Protection Project claims that North Carolina had 13,508
military ballot requests in all of 2008. But, the 2008 absentee ballot file
for North Carolina indicates that there were a total of 9,164 requests from
military voters. To get near the 13,508, one as to combine the military and
overseas civilian requests (the 2008 file reports 13,494 requests). I'd like
to assume that MVPP was not purposely adding together UOCAVA voters in 2008
and then separating out the military voters when making the 2012 comparison,
but given the shaky nature of the report, I am not sure.

See their report here:

http://www.scribd.com/doc/105035437/Bleak-Picture-for-Military-Voters

The Project then claims that there were 1,859 military requests as of, well
we don't know. Footnote 8 of the report describes "phone conversations over
the last 30 days" which would put the information on ballot requests
collected as early as late July given the report's August 28 date. The
accuracy of a report is dubious that compares the number of ballot requests
in this election as of July 30 with the total number requested in 2008.

I've been working with the 2012 North Carolina absentee vote file for a
while. Another confound is that there are non-reporting counties. I really
wish I had the number of non-reporting counties as of August 28, but I
don't. As late as last Friday, there were eleven counties.

Two days ago, I ran a query for NPR making my best comparison of 2008 and
2012, excluding non-reporting counties, excluding non-deliverable addresses
(to account for the MOVE Act changes regarding automatic ballot requests),
and calculating for the same point in time rather than comparing now to the
total number requested in 2008 (adjusting for the fact that the election is
two days later in 2012). There were 5,073 military ballots in 2008 compared
with 4,625 ballots in 2012, a decline of 8.8%. That is certainly much less
than the 86.2% decline claimed in the report. And I caution, some local
election officials who are reporting are slow to update, so the true 2012
number may be higher. Further, there is still time left in the election, so
the final numbers could move in either direction. But if this small decline
persists, I posit an obvious reason: military voters were more excited to
vote for John McCain -- a war hero -- than for Mitt Romney, who spent the
Vietnam War in France on a deferment.

North Carolina overseas civilian requests are up sharply from 2008. Using
the same apple-to-apples comparison, there were 2,760 overseas civilian
ballots requested in 2008 and 4,654 requested in 2012, an increase of 68.6%.
So, if I were to conflate all UOCAVA voters as "military" I could give the
false impression that the number of military ballot requests has increased
over 2008.

The MVPP report is poorly sourced, with in at least one state inflated 2008
numbers to paint a picture of decline even if none existed, that further
inflates a perception of a decline by comparing the total 2008 numbers to
requests perhaps as early as late July, and finally fails to account for
other potential data confounds like non-reporting local jurisdictions. It
has no credibility and should be treated as such.

============
Dr. Michael P. McDonald
Associate Professor
George Mason University
4400 University Drive - 3F4
Fairfax, VA 22030-4444

703-993-4191 (office)
e-mail:  mmcdon at gmu.edu               
web:     http://elections.gmu.edu
twitter: @ElectProject     

_______________________________________________
Law-election mailing list
Law-election at department-lists.uci.edu
http://department-lists.uci.edu/mailman/listinfo/law-election

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://webshare.law.ucla.edu/Listservs/law-election/attachments/20121004/eb980c35/attachment.html>