Back to previous post: Bad morning

Go to Making Light's front page.

Forward to next post: Glad to hear it

Subscribe (via RSS) to this post's comment thread. (What does this mean? Here's a quick introduction.)

November 4, 2004

Why, yes, that is odd
Posted by Teresa at 01:33 PM * 39 comments

From a thoroughly disingenuous article in the Seattle Times:

Larry Harris, a pollster with Mason-Dixon Polling & Research, said the early exit polls overstated the turnout of young voters — people 18 to 29 — and failed to note that Bush received a 4-percentage-point gain in the people older than 60 who voted for him from the 2000 election.

That said, Harris believes that the media and the campaigns misuse exit polls.

“Exit polls are a wonderful advantage in providing context and texture,” he said. “But for trying to make a call, we’ve learned in two elections in a row that they lead to confusion.”

Karyn Barker, a fellow at the American Enterprise Institute, agreed that exit polls are getting a bad rap. It’s not the polls, she says, it’s how the people use them.

“Anyone who follows exit polls should not use them as a prediction of the turnout of certain groups,” she said. “To assume that these numbers were spot-on was foolish. It seems some people, including me, wildly misinterpreted what they meant.”

Some say users of exit-poll information never used to have to worry about error margins and statistical sampling.

“If we go back in history to prior presidential elections, those exit polls were dead on,” said Dennis Simon, a political scientist at Southern Methodist University in Dallas. “Something has changed to make them less dead on.” (Emphasis mine.)
Isn’t it remarkable that all these people (except Dennis Simon) stand ready with explanations for why the exit polls no longer match the voting results?

It’s the of course, everybody knows tone that gets me. I haven’t heard anything quite like it since the day four years ago when James Baker stepped up to the mike and said that of course, everybody knows that machine counts are more accurate than hand counts—which was and is completely untrue. In fact, the reverse is true: everyone knows that hand counts are more accurate.

If this were a movie you were watching, by now you’d be exasperated by characters who took this long to spot the implications. (via)

Furthermore: Jim Macdonald tells me that as he was driving down from New Hampshire on Election Day, he went through a long patch where the only radio station he could pick up had Rush Limbaugh on, so he listened to the show. Jim says Rush was explaining that it’s really not a very good idea to let everybody vote, and that you do much better when the only people who can vote are ones who have “a stake in the system.”

See also, “property qualification.”

As Lydy Nickerson said when I told her about that, “How can anyone not have a stake in the system?”

Addendum: I found this quoted in Red Onion. I don’t know how reliable it is, but it’s certainly interesting:
SoCalDemocrat (xxx posts) Tue Nov-02-04 11:54 PM Original message
Kerry winning Exit Polls - FRAUD LOOKS PROBABLEEdited on Wed Nov-03-04 12:42 AM by SoCalDemocrat
Analysis of the polling data vs actual data and voting systems supports the hypothesis that evoting may be to blame in the discrepancies. Nevada has evoting but with verified receipts. In that state the Exit Polling matches the actual results within .1% accuracy. However for other swing states Bush has unexplainable leads.
I’m still compiling data. Please help me determine what voting methods are being used in swing states and which are evoting without audit trails. Post your data under the individual state responses below. …
When we’re supervising elections overseas to make sure they’re fair and honest, the main mechanism we use is exit polls. On election nights stretching back as far as I can remember, I’ve heard newscasters cite exit polls. Now, all of a sudden, I’m hearing from multiple sources about how wildly unreliable exit polls are.

I’m not buying it.

Also: More links on this subject.

Comments on Why, yes, that is odd:
#1 ::: TH ::: (view all by) ::: November 04, 2004, 01:58 PM:

Too true. But still. Even if the party currently in power has devised a way to stay there regardless of the vote, still there are far too many USians to vote for those guys.

And there is no one in the position to and interested in prosecuting any of the malfeasance.

I'm terribly sorry, but for now, the only thing I hope is that your madness won't rub off too much on us here on the other side of the ocean.

#2 ::: Giacomo ::: (view all by) ::: November 04, 2004, 02:06 PM:

That is a thorough priority for the next two years. State by state (especially swing ones), democrats need to put in place solid voting procedures. Request voter-verifiable paper trails. Request to make impossible to create exclusion lists in the year before the election. Request solid, clear and accountable processes for challengers. Organise quick media-response action groups like Moore did this year. SPEND LESS IN USELESS ADVERTISING AND PAY ACTIVISTS, so they can take time off to do all those "small" operations like keeping accurate lists of who votes what and where...

#3 ::: Mark ::: (view all by) ::: November 04, 2004, 02:20 PM:

All good, Giacomo. Although I'd change one crucial verb: We are way, way past "Request" now. Demand, insist, enforce, do, get. The current bosses don't "request," they act and stay on task and on message until thir goal is achieved, and so should we. To dip into the current lingua freepa, "requesting" is for pussies, or the French.

#4 ::: Reimer Behrends ::: (view all by) ::: November 04, 2004, 02:24 PM:

Hey, does Rush's idea mean that I should be allowed to vote in US elections? Foreigner, living in Michigan, paying income, FICA and Medicare taxes just like your typical American. No taxation without ... eh, nevermind.

(Yes, I'm being facetious.)

#5 ::: Matt ::: (view all by) ::: November 04, 2004, 02:26 PM:

Does anybody have a good reason why you don't divide the electoral college votes for a state based on the proportion of votes each candidate receives within that state? As it stands, you can have a state with 17 votes that gets a 51% return for, say, Mr Monkey-brains, and all 17 votes go to him, when his opponent Mrs Cheesecake-face got 48.5% of the vote. How is that fair?

At least here in the UK we do it with smaller divisions, so although you can win a seat with well under 50% of the votes (depending on how many candidates are standing), it only decides one seat, not seventeen of the things. Strikes me as a simple change which would not make the system perfect, but would improve it considerably - and knock out a lot of this 'swing state' rubbish too.

#6 ::: L.N. Hammer ::: (view all by) ::: November 04, 2004, 02:40 PM:

Matt: It's per the state how to divide the electoral votes. Currently two states do apportion by popular vote. Colorado had a ballot measure to do the same, but it was defeated.


#7 ::: Reimer Behrends ::: (view all by) ::: November 04, 2004, 02:50 PM:

L.N. Hammer: As I understand it, Maine and Nebraska (the two states you are referring to, I believe) pick two electors via a statewide popular vote and then one elector per congressional district based on the popular vote in each such district. So, no, they do not select electors in proportion to the popular vote.

#8 ::: Stefan Jones ::: (view all by) ::: November 04, 2004, 03:16 PM:

That Rush statement needs to be accurately transcribed for future reference.

Rush needs to be found in a hot tub with a live boy, a dead woman, or an inflatable reindeer.

#9 ::: Kait Mackenzie ::: (view all by) ::: November 04, 2004, 03:28 PM:

Why not D. All of the Above? =)

#10 ::: Stefan Jones ::: (view all by) ::: November 04, 2004, 03:35 PM:


E) All of the above, plus Ralph Reed in a Furry otter suit.

#11 ::: Ted ::: (view all by) ::: November 04, 2004, 03:37 PM:

I'd prefer a system where everybody who had a stake in Rush got to vote. The President elect would then be charged with the duty of filling his mouth with communion wafers and burying him.

#12 ::: Merlin Missy ::: (view all by) ::: November 04, 2004, 04:10 PM:

More people need to call their representatives and senators and ask them what they plan on doing about the Diebold / black box voting scandal.

Re Rush: Ooo, I own a house. Does that mean I get to vote more than the neocon woman at work who just rents? What about the value of my house versus the value of the doublewide where my mother lives?

#13 ::: Merlin Missy ::: (view all by) ::: November 04, 2004, 04:17 PM:

(sorry for the doublepost)

Salon's got something, but it's not helpful yet:

#14 ::: Aquila ::: (view all by) ::: November 04, 2004, 04:25 PM:

I have absolutely no idea how accurate and objective this document is. Anyone able to comment?

#15 ::: Christopher Davis ::: (view all by) ::: November 04, 2004, 04:27 PM:

Merlin: even better; the two cities with some of the highest house prices are Boston and San Francisco, those noted conservative bastions!

#16 ::: Jon Sobel ::: (view all by) ::: November 04, 2004, 04:35 PM: is filing huge Freedom of Information Act requests for voting machine records, right down to the Windows Event Viewer logs. Retroactive adjustments of exit polls add more fuel to the conspiracy theory fire. This confirmed skeptic is starting to wonder if this election might really have been stolen.

#17 ::: bryan ::: (view all by) ::: November 04, 2004, 04:46 PM:

"Why not D. All of the Above? =)"

because then it would be a pool.

#18 ::: Richard Bellamy ::: (view all by) ::: November 04, 2004, 04:53 PM:

“If we go back in history to prior presidential elections, those exit polls were dead on,” said Dennis Simon, a political scientist at Southern Methodist University in Dallas. “Something has changed to make them less dead on.”

This is simply untrue. The exit polls we see are the final ones, and the final one this year was about as accurate as the final ones from previous years.

See Mystery Pollster for information on mid-pay exit polls -- that never got a chance to get leaked onto the internet before this year.

#19 ::: Nicole J. LeBoeuf-Little ::: (view all by) ::: November 04, 2004, 04:56 PM:

In defense of Coloradans who voted down the split-electorate amendment (looks shyly at self), there were some real concerns with how it was written. What really decided me was that it proposed to apply retroactively to the 2004 election in which it was decided. Sounded to me like a fishy attempt at using a permanent amendment to our state constitution in order to affect a single election.

For more info, check out Coloradans Against A Really Stupid Idea. I don't agree with all their arguments, but I had enough of my own to vote with them.

That said, I'd love to see a sane proposal to get rid of the winner-take-all system on a nation-wide basis and replace it with something a little more representative. Any suggestions?

#20 ::: Nicole J. LeBoeuf-Little ::: (view all by) ::: November 04, 2004, 05:05 PM:

(Sorry for the double post. Wanted to mention the Boulder Daily Camera editorial opposing the amendment. It's well done, both more informative than "Really Stupid Idea" and more sympathetic towards the good intention of electoral college reform.)

#21 ::: Alison Scott ::: (view all by) ::: November 04, 2004, 05:30 PM:

I put a similar post on black box voting.

I am not suggesting that there has been no tampering. I do not know whether there has been tampering. But.

In the 1992 UK General election, polls shortly in advance of the election, and exit polls, indicated that Labour had won narrowly. In fact, the Tories won the election, with some polls with a '3% margin of error' being far more than that out.

UK ballots are as unriggable as I believe any ballot could be. Crosses are made on slips of paper, which are placed by the voter into locked, sealed, numbered ballot boxes, which are taken by representatives of more than one party to the count, where they are unsealed and counted, again with plenty of people around to check and recount if necessary. So the ballot wasn't rigged.

So, what went wrong with the exit polls? Pollsters and psephologists believe that people voting Tory were so embarrassed about it that they either refused to answer the poll or lied to pollsters.

An article about it here:

For some time afterwards, opinion polls were 'adjusted' to take account of Tories' reluctance to admit how they vote.

#22 ::: Nicholas Rogers ::: (view all by) ::: November 04, 2004, 05:43 PM:

I think the interpretation of "stake in the system" as property rights/ownership is missing the point. The only people with a stake in "their" system are the people who look like them. Who talk like them. And most importantly (and occasionally exclusively) who pray to their god using the words they tell them to use.

Property ownership as the definition of a voter would mean that foreign groups owning significant property could vote and we all know whom that would favor. No, they will always craft a definition that bends towards their "moral values" or some other squishy, irrational abstract concept no matter how unjust or unworkable that definition might be.

Rush and his ideology provide us a wonderful insight on the Republican modus operandi: Exclude, never include…Divide, never unite. The Republic in the word Republican refers to the concept of representative government by the few for the many. Letting everyone vote is simply too much democracy for the control freaks who drive the policy and PR arm of the Religious Right.

They may well try to create a new definition of “eligible voter” but it won’t use anything so concrete as property ownership.

#23 ::: Reimer Behrends ::: (view all by) ::: November 04, 2004, 05:56 PM:

Alison, the effect is more generally known as the Spiral of Silence. It is worth noting, however, that it is a somewhat controversial theory.

#24 ::: Hmph ::: (view all by) ::: November 04, 2004, 06:03 PM:

I'm not trying to advocate property qualifications, or anything like that, but everyone has a stake in the system -- and we aren't letting Canadian or British citizens vote, even though the United States' policies and President affect them.

#25 ::: Dave Bell ::: (view all by) ::: November 04, 2004, 06:21 PM:

The problem with baseing EC votes on congressional districting would be visible, I believe, in Texas.

Just splitting the EC vote in proportion to the statewide vote would avoid that.

I;m not sure we have anybody in UK politics who could get as close to ousting Blair as Kerry did to ousting Bush.

#26 ::: Teresa Nielsen Hayden ::: (view all by) ::: November 04, 2004, 06:25 PM:

There've been a couple of additions to the main post, if you haven't been down that way in a while.

#27 ::: i, squub ::: (view all by) ::: November 04, 2004, 07:05 PM:

This stuff is starting to really freak me out.

#28 ::: sennoma ::: (view all by) ::: November 04, 2004, 07:11 PM:

I'm trying to collect data regarding the exit polls; here's a quick taste from SoCalDemocrat, who's been doing a mighty job:

Kerry is well ahead in Exit Polling in Ohio. We're being screwed.

Male: 51/49 Kerry 47%
Female: 53/47 Kerry 53%

Dem: 91/8 Kerry 38%
Rep: 94/6 Bush 37%
Ind: 60/39 Kerry 24%

Here is exit polling for Florida (3,824,794 votes for Kerry & Bush)

Male: 52/47 Kerry 46%
Female: 52/48 Kerry 54%

Dem: 86/13 Kerry 38%
Rep: 92/7 Bush 39%
Ind: 60/38 Kerry 23%

3,824,794 votes for Kerry & Bush
2065388 Women (54% of total)
1759405 Men (46% of total)

Bush leads male vote by 5% of M = 87970

Kerry leads female vote by 4% of M = 82615

That means Bush is ahead by just 5355 votes in exit polling in FL.

Another odd thing is that there are more Reps then Dems in Florida by 1%, which is not expected. Either there are more voting Republicans in FL than Democrats, a first and not matching known statistics, or more Republicans were exit polled than Democrats. If the exit poll is off by just 1% that's a difference of 382479 more voters who are Democracts.

The results being posted however show Bush ahead 326,000 actual votes. This is simply not possible from the exit polling numbers. Even skewed for a 5% higher Republican vs. Democrat turnout from 2000, it doesn't add up.


Further, here are screencaps showing that CNN altered their exit poll data.

Anyone collecting hard data (pre-adjustment exit poll numbers) please email me what you have. I'll be putting it together over the next couple of days; no doubt someone else will do it faster and better but it can't hurt to have me plodding along behind.

#29 ::: Charlotte Freeman ::: (view all by) ::: November 04, 2004, 07:22 PM:

Okay, I'll jump in with two feet on the tinfoil hat brigade and send everyone over to Eric Francis's astrology blog ( Apparently, the stars all point to a "wounded king" -- the astrology all points to a reign fraught with deceit, and an increasing level of radicalism and an overthrow of the "wounded king". I think we need to push for impeachment. He lied about the WMD, and he stole the election. Impeach the fucker. Be empowered.

#30 ::: David B. ::: (view all by) ::: November 04, 2004, 08:51 PM:

There are many, many reasons to dread another 4 years of Bush and cronies. Astrology is not, however, one of them.

I want blackboxvoting to get those logs. Now. It's quite possible that the results reflect all the fundies coming out for Bush... his margin of three and a half million could be exactly accounted for by Rove's "four million evangelicals" that he predicted would come out this time. But without those logs (and possibly even with them) we have no idea.

#31 ::: Bill Humphries ::: (view all by) ::: November 04, 2004, 11:28 PM:

Ted, you are one with the Funny!

I'd prefer a system where everybody who had a stake in Rush got to vote. The President elect would then be charged with the duty of filling his mouth with communion wafers and burying him.

Thank you.

As one of the organizers of High Stakes, I can get behind that effort. I'll bring the Dust Buster to clean up afterwards.

#32 ::: Jules ::: (view all by) ::: November 05, 2004, 09:54 AM:

It strikes me that there is a possibility that explains exit poll discrepancies that people haven't considered. It's a rather unethical tactic that could be used to manipulate the result of a closely contested election.

Let's keep this entirely theoretical, rather than using any real situations. Say you have two candidates, A and B, both of whom are very close. There are 10,000 electors of which 5,000 would in the normal course of events vote. Without any kind of manipulation, the result of the election would be 2,400 votes to A and 2,600 to B.

Now, lets say 200 people who vote for A are a member of a society that makes a secret decision: they will all vote early in the day and state for exit polls that they voted for B. This has the effect of causing early exit polls to suggest that B has a reasonably wide margin and is going to win easily. People where were marginal in their support of B might now be less inclined to vote on hearing those exit poll results, so B might eventually receive fewer votes because of this. Discourage 101 of those 2,600 voters and A might win.

I don't know if this is a tactic that has ever been used, or even if it would be possible to organise it to work successfully. But it would certainly be possible to try to use, and it might explain some of the discrepancies that have been seen lately.

#33 ::: Jo Walton ::: (view all by) ::: November 05, 2004, 11:23 AM:

Did we just see an election in the "most democratic country on Earth" to elect the "Leader of the Free World", or did we see the accidents of such an election?

I honestly don't know, and that scares me.

Incidentally, for the first time ever in my experience, there were US border people doing an exit check on the train back to Montreal. They didn't do more than glance at my passport, but they were asking Americans around me where they lived, what their jobs were, and how long they intended to stay in Canada, and checking their return tickets. This was ten minutes before the Canadian border, and the Canadian Customs and Immigration inspection there.

#34 ::: Dana ::: (view all by) ::: November 05, 2004, 12:06 PM:

Jules --

Exit poll manipulation has indeed been thrown around as a possibility as to why they were off. But I would like to point out that your theory could just as easily go the other way, in which voters for candidate B inflate exit poll numbers to discourage voters of candidate A by imposing a feeling of inevitability in B's win.

Given all that, the national exit poll data is actually pretty much in line with the outcome of the election. As for why the state exit polls were off, Edison/Mitofsky is still looking into the problem -- at this point, no one really know what happened, or how bad the results really were. But until more information is available, the problems with the exit polls should not be deemed mischief or manipulation or what have you.

#35 ::: Nicole J. LeBoeuf-Little ::: (view all by) ::: November 05, 2004, 12:18 PM:


Incidentally, for the first time ever in my experience, there were US border people doing an exit check on the train back to Montreal. They didn't do more than glance at my passport, but they were asking Americans around me where they lived, what their jobs were, and how long they intended to stay in Canada, and checking their return tickets. This was ten minutes before the Canadian border, and the Canadian Customs and Immigration inspection there.

I've been hearing quips about "The Republic of Gilead", but this anecdote is the closest similarity between real life and The Handmaid's Tale I've heard yet.

My husband and I are already talking about "How long do we stay and fight," and "what's our sign that it's time to get out," and "where do we go if Canada and Mexico aren't far enough?" I'm thinking that passage of the Constitutional Restoration Act in the Senate is our red alert; appointment of new SC justices before said Act can be challenged is an evacuation notice.

But we are going to fight. We're taking a little time to gather a battle plan, and then we're joining the fray. I'm heartened to know that we're in good company here.

#36 ::: Paula Lieberman ::: (view all by) ::: November 05, 2004, 01:05 PM:

I'm thinking about taking a drive (not a colonial farmer) following the path taken by Certain People in April, 1775, to a certain bridge, on the Concord River, past the graves of British soldiers buried where they fell, and lying down and pounding the ground (or the bridge) and CRYING.

That master of mediocrity paraliterate Schmuck and his religiosity fanatics associates, are what all those people gave their lives for over all the past two plus centuries?! I wonder if anyone has set up any protest signs on Lexington Green....

Engraved upon a boulder set in Lexington Green (I am looking at a picture of it)

Line of the Minutemen April 19, 1775


#37 ::: Thel ::: (view all by) ::: November 05, 2004, 01:49 PM:

I deeply resent the way this administration makes me feel like a nutbar conspiracy theorist:

Here are grounds for suspicion. Electronic voting machines figured heavily in the final tabulation of the results in Ohio, Florida, and New Mexico. Moreover, in all three, paper audit trails do not exist.

These states therefore offered the best, safest opportunity for manipulation of the final count.

Question 1: Even if we grant the potential inaccuracy of exit polls, how likely is it that in all three cases the inaccuracy would show a "non-existent" Democratic advantage? Why doesn't the discrepancy ever work in the other direction?

Question 2: Why did problems afflict exit polling in three swing states that have widespread computerized voting with no paper trails?

In other states, the exit polling matched the final results rather well. In Nevada, Illinois, and New Hampshire, computer votes do have paper trails -- and in those instances, the exit polls tracked the final totals.

To recap: In three states with no paper trails, we have exit poll/final tally disagreement. In three states with paper trails, we have exit poll/final tally congruence.


My so-called lesbian life has a short roundup of "sites to start your questioning." Any other good starts, aside from those already posted here?

#38 ::: Paul ::: (view all by) ::: November 05, 2004, 03:36 PM:

Regarding the "stake in the system" comment, this is an expression typically used in the tax debates. I didn't heard the contect of Rush's statement but the argument usually goes like this...

Speaker 1:"People who make under $X a year shouldn't have to pay any taxes"
Speaker 2:"But then they won't have a stake in the system"

The thought is that if you don't have to pay for it, you won't care how much money the government spends on anything, so everyone should be taxed at some level.

The fallicy here is that everyone pays taxes, even if they don't pay income taxes. They still pay personal property taxes and sales taxes.

#39 ::: James D. Macdonald ::: (view all by) ::: November 05, 2004, 08:22 PM:

Rush's argument was, indeed, that the people whose income was so low that they don't pay income tax shouldn't be allowed to vote.

I believe that, regardless of paying income taxes, people of draft age have a "stake in the system." As do those who pay other taxes (sales tax takes a larger percentage of the income of low-income people than it does of high-income people.

Then, how about Rodney Richpygge, who decides to give up his right to vote in return for canceling his income tax? Perhaps he figures that he's buying the lobbyists anyway, so his vote isn't as important as the cash savings.

Welcome to Making Light's comment section. The moderators are Avram Grumer, Teresa & Patrick Nielsen Hayden, and Abi Sutherland. Abi is the moderator most frequently onsite. She's also the kindest. Teresa is the theoretician. Are you feeling lucky?

Comments containing more than seven URLs will be held for approval. If you want to comment on a thread that's been closed, please post to the most recent "Open Thread" discussion.

You can subscribe (via RSS) to this particular comment thread. (If this option is baffling, here's a quick introduction.)

Post a comment.
(Real e-mail addresses and URLs only, please.)

HTML Tags:
<strong>Strong</strong> = Strong
<em>Emphasized</em> = Emphasized
<a href="">Linked text</a> = Linked text

Spelling reference:
Tolkien. Minuscule. Gandhi. Millennium. Delany. Embarrassment. Publishers Weekly. Occurrence. Asimov. Weird. Connoisseur. Accommodate. Hierarchy. Deity. Etiquette. Pharaoh. Teresa. Its. Macdonald. Nielsen Hayden. It's. Fluorosphere. Barack. More here.

(You must preview before posting.)

Dire legal notice
Making Light copyright 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017 by Patrick & Teresa Nielsen Hayden. All rights reserved.