Politics: New Hampshire recount illustrates weakness of electronic voting
I’ve never been sold on the idea of replacing reliable paper-based ballots with electronic voting. While there has always been cheating with paper ballots, it cheating that has to be done by hand, and there usually has to be some sort of conspiracy for it to be successful. Electronic voting can be hacked remotely, or in advance of the elections. And nothing that election judges or observers at the scene can do to prevent it. And detecting it after the fact requires trusting the very same vendors who have a stake in NOT admitting anything wrong happened in the first place.
Consider this entry from The Caucus, the New York Times Politics Blog:
Just three days after Tuesday’s primary election, the New Hampshire Secretary of State’s office says it will conduct a hand recount of the votes in both the Republican and Democratic primaries starting next Wednesday after receiving formal requests from two candidates this week.
On Friday, Albert Howard, an obscure candidate from Ann Arbor, Mich., who appeared on the Republican ballot and received 44 votes in the primary, hand-delivered his recount request and a down payment of $2,000 to the statehouse in Concord, N.H.
And Dennis Kucinich, an Ohio Congressman and presidential candidate, sent a letter to the New Hampshire Secretary of State asking for a recount of the Democratic ballots. Mr. Kucinich’s letter cited “unexplained disparities between hand-counted ballots and machine-counted ballots.â€
The recount is based on concern that in counties when paper ballots were used, Barack Obama seemed to come out ahead. When electronic voting was used (with machines considered very hackable), then Hillary Clinton won.
Meanwhile, Brad Blog recounts statements about the polling data made on the Chris Matthews Show on MSNBC:
MATTHEWS: So what accounts for Hillary Clinton’s victory in New Hampshire? What we don’t know is why the victory is so much different in fact, then the polling ahead of time, including what we call the Exit Polls were telling us. Obama was ahead in those polls by an average of 8 points, and even our own Exit Polls, taken as people came out of voting, showed him ahead. So what’s going on here?
Brad himself comments:
“Why were the polls taken, of people coming out of the booth, so off?,” Matthews tries to ask his guests again and again. And again.
All of them twisted and turned and contorted and grappled and speculated, coming up with every possible unverifiable, backwards-engineered explanation, save for the one that must not be named. The 600 lb. canary in the virtual living room…the fact that no human being has bothered to check what was actually on NH’s vast majority of ballots (80%) which were “counted” by error-prone, hackable Diebold optical-scan machines, all controlled by one bad, horribly irresponsible private company, who has no business being anywhere near a public election…
I don’t really think Hillary Clinton hacked these voting machines. But a recount can only shed some light on what did happen. It needs to be done, and done calmly without all the posturing and court battles that happened in 2000.
UPDATE: Brad Carter points out the history of problems with the machines used in Peoria County.
Billy Dennis is a former reporter and editor at small daily and weekly newspapers in Illinois and Missouri. He comments on media, news and politics. He also owns and operates 
Amazed to see you touting the benefits of paper over computers in this situation……
Obviously Mr. Howard isn’t recounting on his own behalf but this was a very foxy move in order to get a recount for another GOP candidate. Kudos to Mr. Howard for thinking of this!
I happen to know what happened in New Hampshire.
You see, vote totals are comprised of an actual count of those people that cast ballots.
Polls are comprised of small samples of those people that MIGHT cast ballots, or, in the case of exit polls, people who have just cast ballots. The key words being “small sample.”
Now, if I am the pollster and I interview only women, it skews my sample. So maybe I set it up that I will interview 50% men and 50% women. That might be better, but if the actual total of voters is 60% men and 40% women, well, that skews my sample, doesn’t it?
The thing is, it is somewhat easy to separate those two demographic groups. What about income levels, education levels, or religious preferences? Not so easy. If I interview all Mormons, it skews my sample as opposed to a cross section of the population.
Polls can be wrong, sometimes very wrong. They try to make polls sound scientific, but polling science is more of a dark art than science.
The answer is – the polls were just plain wrong.
To Mouse: Exit polls ask the people who have just voted. That is the pertinent sampling. It is used regularly worldwide to determine election fraud. It has been historically accurate. The notable exceptions are in elections still suspect by large portions of the population.
What possible harm could there be in proving the election was honest? I don’t trust computer voting. Period. Scanners are OK by me to get the vote out quickly but count the paper wherever there is a perceived discrepancy.
We need to believe in the system and too many of us do not.