While not being definitive proof of fraud on the part of either the electronic voting machines used in New Hampshire or the human ballot counters, it seems like there is a general discrepancy in the percentage of votes won by senators Clinton and Obama in the primary. In towns that use electronic voting machines, Clinton experienced an advantage, and in towns that hand-count their ballots, Obama experienced an advantage. The other candidates also experienced advantages in one type of voting machine or the other, but the percentages were much closer to zero. I haven't figured out the statistical significance of these numbers, so it could be attributed to random noise. However, I am skeptical of that since the candidates with a higher overall vote count experienced higher discrepancies between electronic and hand-counted ballots, whereas I would think the reverse would hold if it was simply 'background noise'. All the more reason to insist that all electronic voting machines implement a voter-verified paper record of votes.
2008 New Hampshire State Primary Results
Update:
After doing some very simple analysis on the numbers, I've found the following:
Of the 135 towns that used paper ballots, Clinton beat Obama in 47 of them (~35%)
Of the 100 towns that used electronic ballots, Clinton beat Obama in 59 of them (59%)
So, in districts that use electronic ballots, Clinton beat Obama almost twice as often as she did in districts that use paper ballots.
Thursday, January 10, 2008
Subscribe to:
Post Comments (Atom)
2 comments:
Are the towns that count their ballots electronically generally better funded than those that count by hand? It could be that the race is divided along economic lines.
Yeah, that is definitely a possible explanation. I would find it somewhat unexpected, though, given that Obama is said to do better among the higher educated, and Clinton tends to appeal more to the working class. If I had the time, I'd be interested in seeing if there's a stronger correlation between economic status and reported voting preferences than between voting methods and reported voting preferences. There are other demographic statistics might also be interesting to compare. Really, I'd like to see a full statistical analysis of voting preferences, demographics, and voting procedures, but until all of that data is presented in some standard machine readable format, the grunt work is too much for me to tackle.
Kind of a moot point now, though, seeing as Obama has all but clinched the nomination.
I still cringe whenever I hear about electronic voting machines without a voter-verified paper trail, though. Public confidence in election fairness is important if you want high participation from the electorate.
Post a Comment