Slashdot Mirror
Diebold Voter Fraud Rumors in New Hampshire Primaries
Westech writes "Multiple indications of vote fraud are beginning to pop up regarding the New Hampshire primary elections. Roughly 80% of New Hampshire precincts use Diebold machines, while the remaining 20% are hand counted. A Black Box Voting contributor has compiled a chart of results from hand counted precincts vs. results from machine counted precincts. In machine counted precincts, Clinton beat Obama by almost 5%. In hand counted precincts, Obama beat Clinton by over 4%, which closely matches the scientific polls that were conducted leading up to the election.
Another issue is the Republican results from Sutton precinct. The final results showed Ron Paul with 0 votes in Sutton. The next day a Ron Paul supporter came forward claiming that both she and several of her family members had voted for Ron Paul in Sutton. Black Box Voting reports that after being asked about the discrepancy Sutton officials decided that Ron Paul actually received 31 votes in Sutton, but they were left off of the tally sheet due to 'human error.'"
To me the larger issue is the Ron Paul votes that were missing then found again only after the officials were called out on it. This is a very serious problem that can't be refuted or explained away, and I hope it's not overshadowed by the Clinton/Obama issue.
except for the fact that it was reported that someone got 0 (zero) votes.. when voters said they did vote for the guy. which tells you there is a problem.. how many votes for other people didn't get counted? where did the votes go.. did they give the votes to someone else??
showing 0 (zero) just makes it painfuly obvious there is a problem... what we need is to design an effective open system so that there are no errors, or a way that the public at large can be assured that their vote counted.
'...if only "Jumping to a Conclusion" was an event in the Olympics.'
It's also important to note that there's actually a very simple explanation for the results: cities like Clinton.
If you take the cities from TFA (> 5,000 votes, all counted by machine), you get:
That sums up to 97,094 votes (1/3 of the total), of which 42% went for Clinton and 34% Obama. If you restrict to just the largest cities (> 15,000 votes, 13% of total), it's 45% to 31%.
So while it's clear that support for Clinton vs. Obama is correlated with machine-counting vs. hand-counting, it's also clear that both of those are correlated with city size, suggesting a much simpler and rather less nefarious underlying common cause. The tables in TFA don't show that simply because of the highly unbalanced manner in which they split up towns into size classes.
(That being said, of course I'd love to see this be the death knell for vote-counting machines which lack a paper trail. Beats me how anyone ever thought those were acceptable; they may be cheaper than hand-counting, but they simply don't do the same job, making a direct price comparison irrelevant. It's like buying a hammer because it's cheaper than a saw.)
I used to have the unglamorous job of keeping an absolutely horrid, ground-up custom hack job accounting system more or less alive. This thing was written over several years by three or four people who had never met each other in a half dozen wonky relatively dead languages. I had an accounting manager roll into my office in hysterics screaming about how the internal reports and external audits varied by $112...over $250,000,000. This was obviously rounding and not even in error, but even the perceived error was on the order of 0.0000448% -- and that was considered unacceptable, which is a tad absurd when the values in question don't even have that many places. But, we're talking integers here. There is no rounding error.
I mean, come on, the average precinct BARELY record 1000 votes and the biggest don't even hit 3000, yet the voting system for the average high school prom, while equally as complicated, extensive and at risk for fraud, is more secure and less prone to error.
I'm left pretty certain that the only way someone could produce such a system for simple integer tabulation with such comparatively huge error rates is if those errors were in fact deliberate and by design. There seems little other explanation and positively ZERO excuse.
The silly thing is that there are people like you who are arguing here to get rid of the secret ballot, when there's absolutely no need to do so. There's been academic research ongoing for decades in verifiable voting mechanisms, and some of them are truly innovative. Punchscan and Scratch & Vote are two highly visible examples, but there have been scads of other papers on the topic.
Why risk a return to the corruption that occurred when ballots weren't secret when modern technology (not computing, but applied mathematics) provides mechanisms to have our cake and eat it too?
On the other hand, I think it is possible to explain these very strange results without resorting to election fraud. Even so, I do think the current situation warants further scrutiny.
The Independent said there was a 11 point swing between the average of the polls (Obama +8) and the official results (Clinton +3). There are reasons other than fraud for Clinton to beat the polls:
-
Voter complacency after Obama's huge lead in the polls. This would lead more independents to vote in the Republican Primary instead of "wasting" their vote for Obama. Also, some first time voters (like students) may have stayed away from the polls confident that Obama would win easily. This could easily account for 3% of the swing.
- Females deciding to vote for Clinton in the last day. There were two events, both widely publicized by the MST that would have made Clinton more appealing to women. First, the way Edwards came to Obama's defense in the Saturday debate could have made both men appear to be anti-female. Second, the most widely publicized event of the primary was Clinton's teary moment that also might have appealed to females. The exit polls said the late deciders were a wash, they followed the trend of the entire vote. I think the two moments cited above nullified what would have been a swing towards Obama in the late deciders. I'd say this could account for 1 point in the overall 11 point swing.
-
The Bradley Effect where white voters lie to pollsters in bi-racial elections. This is the non-fraud explanation for the 7% discrepancy in the exit polls (Obama +4 vs. Clinton +3). We must give this 7%.
IMO, the discrepancy in the exit polls is the most troubling statistic. If we don't see similar discrepancies (of 5% or more) in primaries in other mostly white states then I think election fraud would be the only possible explanation of the New Hampshire results.We don't see the world as it is, we see it as we are.
-- Anais Nin
If you can call up a web page that shows who you voted for for your own verification, your boss can make you do so to make sure you voted for his candidate.
There's a solution to that problem.
When you vote, the system gives you a single digit number as your verification code. When you go on to the system to verify your vote, it presents you with a list of all the candidates with a single digit number (not the one it gave you earlier) next to each candidate that you didn't vote for and the number that it did give you next to the one you did vote for.
There is then no way for anyone other that you to see who you voted for - all you have to do is lie to your boss about which digit you were given.
This would need to be worked on a little to make it properly foolproof, but it could be done.