South Carolina published their results a while ago, but I'm just getting around to entering all the data and capturing the exit poll information while trying to keep current on other upcoming elections. Here's a recap of the results; Bernie got crushed:
Polling in South Carolina was abysmal; the best performing poll in South Carolina had a greater deviation than the worst performing poll from New Hampshire. The polling very clearly indicated a Clinton victory, by 28%, but polling also failed to assess the scope of her eventual 47% margin. Part of this is due to the timeliness of releases; just two polls were released within the week leading up to the Saturday primary.
Another possible reason for this deviation was the low turnout of just 370,904 voters. Turnout dropped by 30% in comparison to 2008's 532,227 votes [2]; 2016 Hillary would have lost to Obama by 20k votes. John Edwards, the 3rd place finisher in 2008, would have lost to Bernie by just under three thousands votes.
Part of the dropoff can be explained by the participation rate by race; in 2008 [3], the exit polls indicated a 54% Blacks to 46% Whites ratio. In 2016, Blacks accounted for 61% of the electorate and they voted for Hillary by a 72% margin. Polling indicated a 47% margin, so once again, the polling was off.
The rules for our rankings are unchanged since our New Hampshire rankings. We did collect race based demographic information in South Carolina, but that data is not factored into the rankings presented below, mainly due to the sparsity of data. The gender expectation methodology was used.
We've taken each pollster's most recent poll with a 30 day cutoff; polls with blank spaces did not release sufficient information to perform the deviation calculations. These polls are included for reference and are not ranked. Polls toward the top of the table have a lower deviation relative to the final result; a lower deviation implies greater accuracy.
Poll | Topline Deviation | Male Deviation | Female Deviation | Re-Weighted Gender Deviation | Net Deviation |
---|
SC: Monmouth University Polling Institute Feb 14-16, 2016 | 18.42% | | | | |
SC: Anderson Robbins Research, Shaw & Company Research, Fox News Feb 15-17, 2016 | 19.42% | 15% | 25% | 19.1% | 59.1% |
SC: American Research Group Feb 17-18, 2016 | 18.42% | 27% | 15% | 17.68% | 59.68% |
SC: Marist Institute for Public Opinion, NBC News, Wall Street Journal Feb 15-17, 2016 | 19.42% | 20% | 23% | 19.83% | 62.83% |
SC: Clemson University Feb 20-25, 2016 | 21.45% | | | | |
SC: Emerson College Feb 22-24, 2016 | 23.74% | 21.57% | 29.18% | 24.21% | 74.95% |
SC: Selzer & Company, Bloomberg Feb 13-17, 2016 | 25.42% | | | | |
SC: Public Policy Polling Feb 14-15, 2016 | 26.42% | 38% | 19% | 24.41% | 81.41% |
SC: YouGov, CBS News Feb 10-12, 2016 | 28.42% | 45% | 17% | 25.92% | 87.92% |
SC: ORC International, CNN News Feb 10-15, 2016 | 29.42% | | 31% | | |
SC: Gravis Marketing, One America News Network Feb 11-13, 2016 | 29.42% | | | | |
Michigan and Louisiana vote tomorrow; the final projections for Michigan will be posted sometime before the polls close. Pollster rankings for Georgia, Massachusetts, Oklahoma, Texas and Virginia are still being worked on.