2014 general election
2014 once again continued the pattern of a red shift in midterm elections. Despite a massively unpopular Congress, mostly due to the GOP's government shutdown antics, Republicans won massive victories and retook the Senate. House, Senate, and governor's races all experienced significant red shifts. The GOP landslide appeared predictable thanks to pre-election polls that were skewed to the right by the Likely Voter Cutoff Model (LVCM).
Following an off-year in 2012, the red shift in the exit polls again showed up across the board. 19 out of 21 Senate races red shifted (average of 4.1%), as did 20 out 21 gubernatorial races (average of 5.0%). The national aggregate of the House vote experienced a red shift of 2.7%, or nearly 3 million votes. And similar red shifts probably would have appeared in state legislative contests had they been exit polled. In fact, the exit polls likely understated the magnitude of the red shift, since their demographic weighting was drawn from past exit polls that had been adjusted to match past red-shifted elections.
2 Senate races, in Georgia and North Carolina, flipped between the exit polls and official results. However, since the exit polls were already inherently skewed to the right, and some states were split-timezone (allowing the pollsters to adjust the exit poll in accordance with red-shifted official results an hour before releasing them), other states might have flipped too. In particular, this may have been true of Kansas (split-timezone), Kentucky (split-timezone), and Iowa, all of which had small GOP wins in the exit poll that red-shifted to large wins. Up to 5 Senate races may have truly shifted.
3 governor's races, in Wisconsin, Maine, and Kansas, flipped from Democrat in the exit polls to the Republican in the official results. As per the above, the small GOP victories in the exit polls for Georgia, Michigan (split-timezone), and Illinois, plus the tie in Florida (split-timezone), might have really been Democratic wins in the exit polls also. Up to 7 gubernatorial races might have truly shifted.
Fitting with evidence since 2000, it's quite likely that these red shifts are indicative of vote rigging. However, if that's the case, the peculiar pattern of targeting noncompetitive elections needs to be explained. Logically, election riggers would be more likely to target close contests that can be plausibly flipped, and 2006 and 2010 confirmed this. But the 2014 exit polls have relatively few flipped contests, with most of the red shifts only serving to inflate the GOP vote.
Strange as it seems from an election rigging standpoint, there are potential justifications. Padded GOP votes lead to further rightward-skewing of the pre-election and exit polls, helping mask fraud in the future. As an inverse of that, if skewed pre-election polls predict a GOP win that would be unlikely to occur in a clean election, it's necessary to match that GOP skew in the official results.
Electronic vote rigging
- Were the 2014 United States Senatorial and Gubernatorial Elections Manipulated? by Kathy Dopp
- E2014: A Basic Forensic Analysis by Jon Simon
- See the American Legislative Exchange Council page also
- One Person, No Vote: How Voter Suppression Is Destroying Our Democracy by Carol Anderson (2018)