[ol]
[li]Race didn’t matter (much). There is no correlation between voting patterns and county makeup. There is a very small correlation when you look at the precinct level. This is very different from the results of the 2008 Democratic Presidential primaries where Obama did better in areas with more non-white voters.[/li][li]Republicans switching parties is both unlikely due to interest in the Republican primaries and the very high voter turnout. One would expect depressed turnout in the Republican primaries and elevated Democratic turnout. The Democratic turnout was elevated, but Republican turnout wasn’t.[/li][li]There are some fishy results on the Republican side too, such as zero residual voters (i.e. nobody who voted didn’t cast a vote in a race).[/li][li]Analysis of the voting results indicates significant anomalies in the digits of the results. Essentially, statisticians tested the digits in the results and it’s not a likely distribution. This is the type of analysis Nate Silver used to support claims that the Iranian elections were fraudulent in 2009 (http://www.fivethirtyeight.com/2009/06/karroubis-unlucky-7s.html).[/li][/ol]
In short, the results are very fishy and there aren’t any good theories to explain them. Throw in that the very fact that Greene won, in a landslide at that, and it’s not a reach to conclude something is very rotten with the primary results. It may not be limited to the Democratic side, but that’s the one that produced the result that made everybody notice.
Does anybody know what kind of machines and counting systems are they using?
Thanks.
That’s sort of where I’m going, but I’m not specifically concerned about Diebold, more the presence of an verifiable paper trail. Other systems (i.e. butterfly ballot) can lead to strange results.
The fact that the Republican results are fishy make me suspect a bad design rather than conspiracy. Also, you should never attribute to conspiracy what is more easily explained with human error.
For the record, I was supporting the “crossover Republican voters” hypothesis, but mostly because I can’t really think of any other hypotheses that make sense. If the crossover hypothesis doesn’t make sense, either (major contested Republican race for Governor, Republican turnout not depressed, etc.), then I’m stumped. Vote-machine fraud still can’t be ruled out, of course, and would be consistent with everything, but that demands a considerably higher standard of proof than what we’ve got so far.
Let’s hope that the investigative reporter is not yet an extinct breed in South Carolina. I’m glad that the 538 crew is on the case, but there’s a lot they can’t do without shoes on the ground in the state.
Based off the linked 538 article saying the numbers dispell the “crossover Republican voters” theory, and that on ballets, the names are often listed alphabetically by first name… I’m thinking large-scale machine error (unintentional, not fraud).
Even if that is the case ,you should be very worried. It says a lot about machine voting and why we should not allow it. If on the other hand, if it is fraud, it is pretty stupid .
I’d say that a machine problem (in a system with no verifiable paper trail) is a much more serious matter than fraud. An untrustworthy machine system removes the legitimacy not only from the South Carolina primary contests in which it is revealed, but from all results in all states using such systems.
Something I once read, which if true is indefensible and should be unacceptable to any believer in democracy, is that neither independent organizations nor even the relevant governments have been permitted to do proper testing of voting machines! (The reason is that the machine’s hardware and software are proprietary secrets. :dubious: )
That’s well known but the justification for it on business terms, says the governments and the peoples right to make sure our elections are fair and honest , is subservient to the right of a corporation to make money. I can not imagine allowing that kind of power to be centered in company whose interests are not in a clean elections.
This here graph, taken from the link in Jonathan Chance’s post, is the smoking gun if there ever was one. It demonstrates that there is no link between racial makeup of counties and Greene’s share of the vote. Typically, when you run stats analysis of SC statewide races, you’re looking at a coefficient of determination of about 0.70. Here, it’s 0.04.
http://www.youtube.com/watch?v=-s9PkuiIw2Q&NR=1 It is even worse than you might think. One of the voting machine companies has been bought by a Venezuelan company. So they are monitoring and safe guarding our elections.
Good point. Something very, very strange is afoot.
So…is Greene fully cognizant of what’s going on, or do you think he’s an unwitting dupe? If the former, he’s certainly playing dumb well enough to convince most folks of the latter.
The lack of racial correlation can be easily enough explained by nobody having a clue what race Greene was. But that still leaves his victory to be explained, of course.
Indeed. Reading through the 538 articles, I was most struck by Rawls’ campaign manager’s comments. He seemed totally baffled by the whole situation…to the point that he stated that he periodically checked in on what Greene was up to during his “campaign” and found he was doing nothing. He didn’t even attend a Democratic party rally in his own city.
It’s too bad that there wasn’t more exit polling done here, probably because it was a low-key primary that wasn’t attracting much attention. It would have been interesting to hear exactly what Greene voters cited as motivation…or, more to the point, if there were any Greene voters.
Every person I have ever met in SC whose last name was Green or Greene was either Black (mostly) or Jewish (lesserly). IJust sayin’. Take that as you may. I didn’t know before the night of the election, but I would have had a clue.
