Another exit poll hint

Sure enough, some exit poll numbers are now making their way around the web. The only ones (out of context) that look to be remotely statistically meaningful suggest Santorum (Pa) and DeWine (Ohio) are in big trouble. But we kinda knew that.


Indiana and Kentucky tea leaves

The five early U.S. House races may help us calibrate the rest of the election, but don't get too excited about the wrong thing. Hostettler will lose IN-8. He's considered one of the 13 most likely to lose (Dems need 15 to turn control). The other two Indiana races are IN-6 and IN-9. They are part of a tossup group that gets the Dems from 13 to 27. KY-3 (Anne Northrup) is part of that same group. KY 4 is widely considered to be more likely to stay Republican. If Lucas beats Davis, it would be one harbinger of a 30-seat sweep for the Dems.

Another caution, any one seat can defy a trend. An occasional safe seat will surprise us, but the overall groupings are good measures of trends.


Go team! Don't win too much?

The Boston Globe's James Pindell suggests John Edwards may have a reason to root against a big Democratic pickup in the New Hampshire Senate. "That’s because of the eight Democratic incumbents, four are hard-core supporters. The more Democratic state senators there are, the more there are to go around and endorse his opponents."

I'll be happy to post a comment from an Edwards supporter out there... 


One exit poll hint (sorta)

According to ABC, "Preliminary exit poll results indicate that nearly six in 10 voters today disapprove of the way President Bush is handling his job."

Their wording suggests something like 58% disapproval and about 42% approval. That would actually be slightly better for Bush than recent polls. It's hard to try and read anything into the number, but it is suggestive that voting is likely similar to recent breakdowns, meaning only that the close races will be close.


Up next: abortion

While you're waiting for the polls to close and numbers to come in, you might as well get up to speed on the major abortion cases the U.S. Supreme Court will hear tomorrow. Lyle Denniston of Scotus Blog says of the timing:

"It is probably not coincidental that the case was scheduled for argument after election day. The nation as a whole, and the opposing sides in the abortion wars, will be watching the outcome of a referendum Tuesday in South Dakota on a law that the state legislature enacted with the explicit aim of putting the future of Roe v. Wade to a test; first at the ballot box, ultimately in the courts. The Court no doubt was aware of that, and sensitive to it."

He has much more on the issues in the case here. Audio and transcripts will be available at the court's website tomorrow.


Turnout lagging?

After receiving reports from polling places across the state that turnout was lower than anticipated, the Democrats today launched door-to-door poll teams to encourage voters to go to the polls.

"This is something we were prepared for, and we were ready to implement, because turnout is important and every vote is going to count," said Kathleen Strand, a spokeswoman for the state Democratic Party. The mobilization of the poll teams came in addition to scheduled Election Day get-out-the-vote activity from the Democrats, including statewide phone banks from the party's 10 offices.

As of mid-afternoon, turnout was lower than expected in the Seacoast, Manchester and Keene areas in particular, Strand said. "These claims of record turnout -- that's now what we're seeing in the numbers coming back this afternoon," she said.

- Eric Moskowitz


Why no exit poll gossip?

Most election-day afternoons of late, the internet has been full of reports of what the exit polls are saying. Not so today -- at least until 5 p.m., when the people with the numbers get let out of their cell. No, really. Howard Kurtz reports:

"The biggest behind-the-scenes change in network coverage involves what has been dubbed the Quarantine Room. Determined to avoid a rerun of recent years, when its exit polls leaked out by early afternoon to the Drudge Report, Slate and other Web sites, a media consortium is allowing two people from each of the networks and the Associated Press entree to a windowless room in New York. All cellphones, laptops and BlackBerrys will be confiscated. The designated staffers will pore over the exit polls but will not be allowed to communicate with their offices until 5 p.m."

For a good discussion of how exit polls are done, how accurate they can be and how inaccurately their information can be used, see here.

UPDATE: The official Republican Party line on exit polls is here.


Missing the micro-target

This get-out-the-vote business is less than a perfect science. We just heard about a recorded phone message from the wife of state Sen. Bob Flanders asking a Bow resident to support her husband. The thing is, Bow isn't in Flanders's district. (Guess our map could have helped.) No telling how the Flanders pitch will affect the Gatsas-Backus race.

UPDATE: A reader suggests this sort of glitch could happen if a former resident of the district moved out of district but took his phone number with him. Me: Yet another way to confuse pollsters!


Another prediction

Robert Novak predicts in his weekly newsletter that Democrats will gain 19 House seats, two Senate seats and five governorships. He predicts Paul Hodes will narrowly defeat Charlie Bass.

A piece of his analysis: "For months, many polls have shown Democrats doing better among the carefully screened sample of 'likely' voters than among the great unwashed mass of 'registered' voters. Historically, it has been the opposite: Republicans have performed better when the likely non-voters are excluded. Polls are now reflecting a return to that historical normalcy -- Republicans perform better among 'likely' voters. This is a small but noteworthy sign that the GOP base is coming home and will vote rather than sit it out. If Republicans were to lose their historic advantage of their registered voters' turning out more reliably than those of Democrats, they would likely suffer a disastrous loss of more than 30 seats. They seem to have avoided the tsunami that everyone had been talking about."


Be careful what you wish for

Mickey Kaus makes the case for why the most likely outcome of the national elections is the one that will produce the result most likely not to accomplish what the most motivated voters most want to accomplish.

Shorter Kaus: Since we can't fire Bush, he'd like the Senate to flip, not the House. Alas, it looks like the House will flip, not the Senate. (And yet he voted for the Dem in his House race.)


About those "undecided" voters

Undecideds -- can't rely on 'em, can't exclude 'em from your polls altogether.

The final WMUR/UNH tracking poll will report that 11 percent of likely 1st Congressional District voters and 10 percent of likely 2nd District voters remain undecided. Is that possible? Two days before the election, were they really unsure which candidate they would vote for? Or were they just unsure whether they would vote?

You will often hear that undecideds "break" to the challenger. This theory tells you that if the last poll has 10 percent undecided, add 6 or 7 to the challenger and 4 or 3 to the incumbent. I'm not so sure that's a great model this time around.

Clearly the most energized voters are Democrats, and the issues over which they are most energized are Iraq and President Bush. If those issues have yet to capture your attention to the point where you know you're going to vote, isn't it too late for these to become the deciding factors?

Look at these numbers from last night's WMUR/UNH release: In the 1st CD, Jeb Bradley led Carol Shea-Porter 47-43 (10 percent undecided) among those saying they will definitely vote and 61-23 (16 percent undecided) among those saying they will probably vote. In the 2nd CD, Paul Hodes led Charlie Bass 50-39 (9 undecided) among definite voters and trailed 40-27 (26 undecided) among probable voters. My read on these numbers is that some of these "probable voters" probably won't vote -- and as turnout expands, the incumbents are the ones who stand to do better.

UPDATE: The final UNH numbers are here. Bradley leads Shea-Porter, 49-40 (+/-5.3). Hodes leads Bass, 48-39 (+/-5.5). Survey director Andy Smith then makes these predictions: Bradley, 53-47. Hodes, 54-43 (with 3% for Libertarian Ken Blevens).

The final splits for definitely/probably vote are 46-46 and 57-26 for Bradley and 53-36 Hodes/46-34 Bass. Also, Smith says he looked back at 2002 and 2004 and found that 10 percent undecided in his final poll is consistent with those years' results.


The end is near. Or not.

I'm reliably informed by "Republican insiders" that the world is coming to an end. Woe is them, and nothing can be done. We should all accept fate and revel in a new order.

On the other hand, I am informed by bright-eyed wise men that every poll ever taken is wrong and misleading, designed only to alienate us from each other. Except the good one, of course.


Betting on polls

Here's a worthwhile post at the pollster.com blog exploring the Big Question: Do polls really tell you who will win? The answer, not surprisingly, is yes and no. But there is a strong improvement in a poll's predictive value as its reported margin increases:

"By the time we see a 10 point lead in the poll for the Dem, about 90% of the Dems win. When we see a 10 point margin for the Rep, about 90% of Reps win. . . . Polls rarely miss the winner with a 10 point lead, but they DO miss it 10% of the time.

"A 5 point lead, on the other hand, turns out to be right only about 60-65% of the time. So bet on a candidate with a 5 point lead, but don't give odds. And for 1 or 2 point leads (as in some of our closer races tomorrow) the polls are only barely better than 50% right in picking the winner. That should be a sobering thought to those enthused by a narrow lead in the polls. Quite a few of those 'leaders' will lose. Of course, an equal proportion of those trailing in the polls will win."


Read up, go vote and see who wins

We're going to be live-blogging the election Tuesday night. Just refresh this page for continuous updates and analysis.

For now, we're keeping this post on top so you can get up to speed. The Monitor's 2006 Voters' Guide is online here. It contains information on the races for governor, Executive Council, Congress, state Senate and state House; the two constitutional amendments on the ballot; the Concord city council and school board races; and the races for Merrimack County sheriff, register of deeds, register of probate and treasurer and Belknap County attorney.

Here are links to the Monitor editorial board's endorsements:

  • Kass Ardinger, John Callewaert and Betty Hoadley for Concord School Board
  • Matt Newland for Concord City Council
  • Republican Scott Hillard for Merrimack County sheriff
  • No on Question 1
  • No on Question 2
  • Democrats Sylvia Larsen, Kathy Sgambati, Harold Janeway and Bob Backus for state Senate
  • Democrat John Lynch for governor
  • Republican Jeb Bradley for Congress, 1st District
  • Democrat Paul Hodes for Congress, 2nd District

If you'd like to read some opposing viewpoints, here are some places to start:

  • Yes on Question 1
  • Yes on Question 2
  • Various Republicans for state Senate here and here
  • Republican Jim Coburn for governor
  • Democrat Carol Shea-Porter for Congress
  • Republican Charlie Bass for Congress

Capital Beat extra: He loves what we've done with our ballot

Out of the blue, we got a phone call from Jon Krosnick, the Stanford professor whose expert-witness testimony a few years back convinced the New Hampshire courts, in a case that challenged the state's ballot-order laws, that the primacy effect -- the advantage of going first on the ballot -- is real.

That case reached the state Supreme Court, which ruled last summer that it was unconstitutional for New Hampshire to give the leadoff column on the ballot to the party that had amassed the most votes in the previous election -- a law that had kept Republicans first for four decades -- or to name candidates in alphabetical order within party lists. After considerable hand-wringing by the secretary of state and a rare special session of the Legislature, the state adopted a system that will give Democrats, Republicans and independents or third-party candidates an equal chance to appear in the first column on tomorrow's ballot. The ballot will also list candidates in alphabetical order starting with a randomly drawn letter (K), in a K-Z, A-J fashion.

Krosnick applauded the changes. "You're making history in New Hampshire, which is great," he said. "I wish so many other states would get smart."

In most states, he said, legal challenges have been unsuccessful. The difference in New Hampshire was partly due to the state's constitution, which confers the equal right to be elected as well as the equal right to vote. These cases have faltered in the past over the question of whose rights are being violated, those of the voters or the candidates, Krosnick said. And recent advances in scholarship have made the case for the primacy effect considerably stronger, he said.

"New Hampshire ought to be a model and ought to be an inspiration for the rest of the country," he said.

- Eric Moskowitz

UPDATE: Then again, maybe the new ballot format won't go over so well.


« first‹ previous171819202122232425next ›last »