Saturday, January 12, 2008

New Hampshire: Which wine goes with chili dogs? (Added note)

Salon writer Farhad Manjoo -- whom my ladyfriend affectionately calls Moojuice -- is having another go at the election integrity advocates. This time he's targeting the folks who think something odd occurred in New Hampshire.

In a sense, both his words and mine are irrelevant. Kucinich has called for a recount, a paper trail exists, and New Hampshire law disallows machine counters. Unless a chain-of-custody question arises, the new count should settle the matter. To his credit, Manjoo supports the Kucinich effort.

But his argument is still worth questioning. Why did the hand-count counties go for Obama while the machine-count areas went for Clinton? "Those places simply vote differently," or so sayeth Moojuice. He draws figures from this extremely useful page to make his point: Small towns do not use optical scannners, and small town folk like Obama.

Hmm. Doesn't that presumption conflict with this?
In the New Hampshire primary, Obama did better among Democratic voters who make over $50,000; Hillary did better with the under $50,000 set. He’s the “wine track” candidate, Hillary is the “beer track” candidate.
In New Hampshire, do the upscale chablis imbibers prefer to live in the stix? That idea seems counter-intuitive, to say the least. We need data.

(To read the rest, click "Permalink" below)

So let's take a few Obama-leaning small towns listed here, and then let's look at some income data as derived from this page and various town websites. This is per capita income:

Brookline: $29,272
Caanan: $20,515
Campton: $20,189
Chesterfield: $25,051
Chichester: $24,115
Claremont: $20,267
Colebrook: $18,390
Columbia: $16,859
Concord: $21,976

Well, an obvious pattern has emerged, and we have not even gotten out of the Cs.

Of course, there are differing ways to look at income. In Canaan, for example, per capita income is $20,515; the average male earner makes $32,946, while the average female earner makes $25,000 and the average household income is $43220. In fact, there is only one town in all of New Hampshire (New Castle) where the per capita income is above $50,000. Folks there did indeed prefer Obama to Clinton, 190 votes to 90. In Manchester -- the state's biggest city, which broke for Hillary in a big way -- the per capita income is $21,244.

So, yeah, we can quibble over the best way to interpret the figures given above. Still, I think we can fairly say that the hand-count towns -- the ones that favored Obama -- are not noticeably more affluent than the machine-count towns. One would therefore think that Obama would do at least as well in the larger cities, where folks are more likely to land jobs that pay 50 thou a year.

Yet he didn't. At least, not officially.

To recap: The data tell us that those snooty Cabernet Sauvignon connoisseurs really dig Obama, while Clinton plays to the Bud guzzlers. But bring up the dreaded specter of election fraud, and suddenly the whole story changes: Obama appeals to the Miller High Life slobberinos, while Clintonites order the Zin.

Well...which is it?

One last thing: Moojuice can't resist taking a dig at exit polling -- which, as you know, is our only check on vote-rigging.
Regular readers might know that I've long been skeptical of efforts to use exit polls -- surveys of voters as they leave voting booths -- as a forensic tool to detect fraud. Academic election experts and pollsters (including the surveyors who carry out the exit polls) say that exit polls, at least as they're practiced in America, are not precise enough to catch fraud in close races.
And how do we know that any given race is close? Because of the final tallies, and because of the pre-election polls. But the final tallies constitute the very point in question, and the folks who dis the accuracy of exit polls will always really dis the accuracy of pre-election polls. Circular reasoning.

Like it or not, international election observers (such as the OSCE) do use exit poll data as an indication of possible vote fraud. (See the previous story here.) I hope Moojuice calls up the good folks at the OSCE to tell them that they are doing their job all wrong.

Added note: Egads. Problems with the above argument have suddenly become clear to me: The argument does not take into consideration the percentage of the NH population that is affluent. While it still seems counterintuitive to find the wealthy in the tinier burgs, a small town in NH may not resemble a hardscrabble community in, say, Georgia. And factors other than income may have solidified support for Obama in the hand-count areas.

Still, the swing to Hillary in Manchester does look suspiciously huge to me. And if you read Moojuice carefully, he never suggests any reason as to why the hand-count areas would have differing political views -- and he never employs the word "maybe" or any similar qualifier (as Brad does). He simply pronounces It Is So.

We shall see.
Just as not all polls have sound methodologies, so too can exit polling be done badly.

It seems multiple parties were doing exit polling, rather than having a single source for it and sharing out those singular results. (Chris Matthews mentioned 'our' (presumably NBC or CNBC) exit polls. 'Donahue' (Phil?) was on Hardball, and he stated from what he saw himself that the exit pollsters appeared far from organized, very young, and about as professional as magazine subscription solicitors.

I asked before and didn't see an answer, but presumably, exit polling massages its raw numbers in accordance with a model of the electorate.

However, some of the professionals discussing these results have said that with a 57% turnout for the primary, it resembled more of a general election turnout than a primary turnout, and that their modeling was simply inadequate (i.e., wrong) for such a case.

Certainly, women made up more of the turnout than expected, and with the mild weather (61 F, from what I remember was said), probably the elderly turnout was better as well.

At least 1 other person is looking for a recount. These excerts are from the Ann Arbor News paper edition, Friday, January 11, 2008.

[Albert] Howard, among 42 lesser-known candidates who paid the $1,000 fee to appear on that state's primary ballot, returned to New Hampshire on Thursday to officially challenge the results. He received 44 votes, but insists he had more that seemingly disappeared and plans to demand a recount.
The 41-year-old father of eight said he was watching C-SPAN and was pleasantly surprised to see a tally of 50 votes with just 12 percent of returns in early Thursday night. About an hour later his total went as high 187 votes....
He said New Hampshire Secretary of State William Gardner could not explain what happened and told him that a recount was the only way to confirm the final tally.

Here's his webpage about the recount:

Very odd video on there about how the numbers for Ron Paul are wildly inaccurate.

Disclosure: I want neither of these guys as president. I'm only bringing this up because it's about voting numbers not adding up.

Post a Comment

<< Home

This page is 

powered by Blogger. 

Isn't yours?