A quick analysis of raw data for the 2008 and 2012 primaries from the North Carolina State Board of Elections on absentee voting for the primaries shows significant improvement in absentee voters getting their mail-in ballots successfully returned for counting. Comparing 2008 and 2012 primary election turnout showed substantial improvement in voters successfully casting ballots. Civilian ballot success went up from 74.4% to 83.9%, military from 26.8% to 31.4%, and overseas from 44.1% to 45.1%. Those changes are probably largely because of absentee ballot reforms in the 2009 and 2910 session of the NC General Assembly that simplified the process with special aim at military and overseas civilian voters.
The reforms as they applied to the 2012 primary (i) cut the number of witnesses required for absentee voters from 2 to 1, and (ii) provided that rather than having to be received by the day BEFORE the election, civilian absentee ballots must be postmarked by election day and received by the Friday after the election, and military and overseas ballots have to be received by the day before the county canvass (making this year’s deadline 6 days after the primary). For the 2012 general election ballots will go out 60 days before the election, 10 days earlier than in 2008.
Here are some comparables with the numbers showing ballot applications received and ballots successfully cast:
===========
CIVILIAN ballots from the 50 states + DC
2008 primary 23,576 of 31,699 74.4% success rate
2012 primary civilian 20,312 of 24,199 83.9% success rate, but of those 1,834 were received during the Tuesday-Friday extension, if the old deadline had applied the success rate would have been 76.4%.
Thus for civilians there was a slight improvement by the old return deadline, but the extended period substantially increased the chance a voter’s ballot was successfully cast.
===========
MILITARY
2008 primary 1,043 of 3,892, a 26.8% success rate
2012 primary 584 of 1,858, a 31.4% success rate, but of those 89 were received during the six-day extension, if the old deadline had applied the success rate would have been 26.6%.
Of note is the sharp drop in military applications between 2008 and 2012, possibly because of a reduction in overseas and other deployments during the 4-year interval.
While the overall success rate in the 2012 primary for military voters was 31.4%, 581 ballots were mailed out automatically to voters who had voted by mail in 2010, and the success rate there was 19.4%. Probably accounting for the low return rate is the inability go get mail delivered to military personnel with 2 year old addresses and discharges of military service perssonel in the two-year gap. Of military voters who applied for ballots in 2012, 36.9% were successful in getting their ballots counted.
=============
OVERSEAS CIVILIAN
2008 primary 450 of 1,021, a 44.1% success rate
2012 primary 661 of 1,466, a 45.1% success rate, but of those 94 were received during the six-day extension, if the old deadline had applied the success rate would have been 38.7%.
==============
Interesting were some of the reasons that the 3,872 civilian ballots that were sent out were not counted. While 3,014 were mailed out and simply not returned, here are some of the noted reasons boards of elections noted that the ballots did not count:
221 cancelled at request of voter (maybe the voter decided to vote in person?)
82 received by May 9-11 but not postmarked by May 8.
58 returned undeliverable by the USPS
148 ballot envelope not signed by voter
56 ballot envelope not signed by witness
288 spoiled ballots (returned by voter with a request for a replacement ballot)
50 rejected by county board of elections (reasons not specified)
23 returned by the voter with notation that they did not want to vote
Posted by gercohen 
Drafting Musings 