November 2, 2006

A Tale of Two Turnouts

Hatched by Dafydd

Real Clear Politics links to Larry Sabato's Crystal Ball, which has just released what might be Sabato's final prediction for the 2006 midterm elections. (He could release a new one at any time, of course; it's a web page, not a print magazine!) Today, Professor Sabato predicted that the Democrats will pick up 6 seats in the Senate and 27 seats in the House, seizing control of both bodies (fairly strong control, in the case of the House).

I just went through the 50 most vulnerable races in the House and the 13 most vulnerable races in the Senate, per the RCP election pages, averaging all post-October 15th polls in every race for which there were polls, and an interesting pattern emerged:

  • In the House, Democrats are currently ahead in 29 of the 45 races in Republican-held districts; Sabato predicted they would pick up 27 of those 29, or 93%;
  • In the Senate, Democrats are currently ahead in the polls in 6 of the 8 Republican-held states; Sabato predicted the Democrats would pick up all six of these races -- 100%.

These "leads" include quite a few in the House where the lead is 1%, 2%, or 3%; and in half of the Republican Senate seats where a Democrat leads, the lead is less than 3%... in fact, it's only 1.4% in MO and 0.8% in VA. But even so, Larry Sabato predicts that Democrats will win nearly all of these -- along with holding every seat of their own: the Senate seats in New Jersey and Maryland and the two House seats in Georgia and Illinois.

In addition, three House races in New York, where Dems are ahead in GOP-held districts, include some very strange results; in NY019, NY-20, and NY-25, the only public polls putting Democrats ahead are those pesky RT Strategies/CD polls... but in each case, RT STrategies/CD gives a whopping lead to the Democrat... far more than even Democratic polls do!

  • In NY-20, two Democratic polls put Kirsten Gillibrand ahead by 2.0 and by 3.0; a Siena College poll puts Republican John Sweeny ahead by 14 points... but the last two RT Strategies/CD polls put Gillibrand ahead by 12 and 13 points, more than six times her lead in the partisan Democratic polls;
  • In NY-19, a Democratic poll actually puts Republican Sue Kelly ahead by 2 points; but two RT Strategies/CD polls put Democrat John Hall ahead by 9 points and 2 points [this is corrected, per commenter Pete; I had switched the two names... but the point is accurate: the Democratic poll has the Republican ahead -- but the RT Strategies/CD poll has the Democrat ahead];
  • And in NY-25, Democratic polls also put the Republican ahead by 2 -- while two RT Strategies/CD has the Democrat ahead by 8 and 9.

(This is one of several reasons why I have lost nearly all faith in the RT Strategies/CD poll... it's just wacky.)

Knock these three off the charts -- not even Sabato himself predicts a Democratic victory in any of them -- and we're left with the conclusion that Sabato, in order to get his 27, must be predicting Democratic victory in at least one Republican district where the Republican is currently leading.

What can explain this? Simple: Larry Sabato is actually predicting a Democratic wave that will wash nearly every close race into their pockets. In other words, he agrees that the pollsters have bad turnout models: but Sabato believes they're being too biased towards Republicans... because ordinarily, a series of toss-up races is not all won by one party.

We can only get to 27 Democratic captures in the House and 6 in the Senate if Professor Larry Sabato believes that Democrats are elated and will show up in record numbers, while Republicans are depressed and will stay home in droves. His mental turnout model has the Democrats, not the Republicans, with the better ground game, such that they win races where they are only 1 or 2 points up now, and even win races where they are 2 or 3 points behind.

By contrast, as I have cautioned many times, the Big Lizards mental turnout model is just the opposite: I see Republicans not as more enthusiastic than Democrats, but more enthusiastic than the pollsters' turnout models, meaning that the polls are slightly biased towards Democrats. Similarly, I believe the GOP GOTV is superior to the Democratic GOTV, and the former will do a better job of turning out party faithful. I believe these two points will lead to a general 3% - 4% "unexpected" bump for the Republicans (which will be portrayed by pollsters as an "election-eve rally")... and at the moment, looking at the polls of today, that would lead to a Republican loss of about 12 seats in the House and 3 seats in the Senate.

All we can say for sure at this point is that Larry and I cannot both be right: one of us is -- or both of us are -- completely, utterly, catastrophically wrong.

Thus, this race has come down to a tale of two turnouts: the one believed by Karl Rove, Ken Mehlman, Hugh Hewitt, Power Line, Big Lizards, and many other fast & furious new-media commentators... and the one believed by Howard Dean, James Carville (in public statements, at least), Fred Barnes, Mort Kondracke, Larry Sabato, and many other "Beltway Brahmins." In short, the one where Democrats have a mild breeze behind them, and the one in which it's more like a hurricane.

Am I hedging? Sure: I could be completely wrong about the turnout. Maybe Sabato (who is even wilder in his pro-Democrat predictions than Mort Kondracke) is totally right; maybe having buttered my bread, I will have to lie in it. But I'm not backing down just because a bunch of relentlessly conventional critics parrot the relentlessly conventional wisdom.

I'll see you all after November 7th -- where I will either make a triumphalist pest of myself, strutting about the board as if I were High Potentate of Next-to-Nothing and Windbag Extraordinaire; or at the very least, I'll go down swinging and take a few of the bad guys with me.

Heck, I've always wanted a honor guard on the road to Hell.

Hatched by Dafydd on this day, November 2, 2006, at the time of 5:26 PM

Trackback Pings

TrackBack URL for this hissing: http://biglizards.net/mt3.36/earendiltrack.cgi/1419

Listed below are links to weblogs that reference A Tale of Two Turnouts:

» Sabato's Crystal Ball 2006: THE PREDICTIONS from The Moderate Voice


Larry Sabato and his staff predict a major defeat for Republicans at all levels of government.

[Read More]

Tracked on November 2, 2006 7:26 PM

Comments

The following hissed in response by: Fritz

Ah Dafydd, you are my favorite "High Potentate of Next-to-Nothing," and "Windbag Extraordinaire." None-the-less, while I think the repubs will hold the senate, losing three seats, possibly four, I would not be surprised to see the house go Democratic by a narrow margine. I don't see how the Dems can pick up over twenty at the most, but I wouldn't be shocked if they managed to win only twelve. My guess is fifteen or sixteen, enough to change control. So now I'm on record also.

The above hissed in response by: Fritz [TypeKey Profile Page] at November 2, 2006 5:40 PM

The following hissed in response by: Big D

No question - you are right.

When in the last 30 years have pollsters not been biased toward the Democratic party? I'm not faulting the pollsters - I think they are doing the job as best they can. However there is significant bias between what someone says in a poll (or even who answers a poll) and who actually gets off their ass and votes.

Maybe the polls are consistently wrong because Democrats are consistently narcissistic and lazy.

The above hissed in response by: Big D [TypeKey Profile Page] at November 2, 2006 5:42 PM

The following hissed in response by: Dafydd ab Hugh

Big D:

When in the last 30 years have pollsters not been biased toward the Democratic party?

Well, actually two years ago: the pollsters did pretty well in the 2004 elections... as did Larry Sabato. As did I, though I wasn't blogging then; I was actually more accurate than Sabato -- I correctly guessed one more Senate race going to the Republicans than did he.

In the Senate, I predicted a GOP pick-up of 3, Sabato predicted 2: it was actually 4. And in the House, we both predicted a GOP gain of 3, which was bang on. (I won $200 in bets that year.)

Of course, he also infamously predicted that the electoral college would end up tied at 269, with the election being decided by the House of Representatives (one vote per state delegation, so of course Bush would have won)... and it actually turned out to be 286 to 251 in favor of President Bush. So there!

Sabato predicted dead even in the popular tally; I predicted Bush by 4%; it was actually Bush by 2.4%, so we were equally off there... but in opposite directions.

Note that when Larry Sabato is wrong, it is almost always in favor of the Democrats. I don't know whether he is a Republican or a Democrat now, but he got his political start working for populist Democrat Henry Howell, Virginia statehouse pol.

Of course, Sabato was only 15 then, and he could have shifted to the Republican Party, if he were a conservative. I just don't know.

Dafydd

The above hissed in response by: Dafydd ab Hugh [TypeKey Profile Page] at November 2, 2006 6:12 PM

The following hissed in response by: Maetenloch

Dafydd, I'm with you. While I have a lot of respect for Sabato and the other pundits, sometimes they get so caught up in the inside game that they fail to do a sanity check on their predictions. My best guess is that the GOP will lose the house but barely keep the senate. If you look at Tradesports, that's pretty much what other people are betting as well. GOP control of the senate is trading at 67 while control of the house is down to 25. After reading James Surowiecki's 'The Wisdom of Crowds', I'm willing to bet money that on Nov. 6th the closing prices at tradesports will be closer to the actual results than what Sabato et. al predict.

The above hissed in response by: Maetenloch [TypeKey Profile Page] at November 2, 2006 7:18 PM

The following hissed in response by: Watchman

I think Sabato's smear job on Allen regarding the "n-word" pretty much reveals his partisan leanings. The truth is I don't think any of us really know what's going on. The polls are so screwy that it's hard to rely on them...even when they say what I want them to say (which isn't that often lately).

It all comes down to turnout, and in places like Maryland and even Michigan, we could get what the polls say, or we could get big upsets. Nobody expected Georgia in 02 after all.

I still think that it's (just barely) possible for the Repubs to end the night losing a net of only 1 in the Senate and still holding the House. Sure that's a best-case scenario, but I can at least make an argument for it with a straight face--especially if there's none of my money on the table!

The above hissed in response by: Watchman [TypeKey Profile Page] at November 2, 2006 7:44 PM

The following hissed in response by: Towering Barbarian

Polls have 2 functions. The more respectable one is as a diagnosis tool which doesn't always go too well even in the Presidential elections (Landon/Roosevelt and Dewey/Truman for the most spectacular examples). The other use is as propaganda (Energize the faithful and daunt those who would vote for your opponents). In the end the voters vote as the voters vote. My best guess is that at worst this will be a typical election of the sort that happens when we're 2 years into a President's term and at best the Donkey Party will be surprised by how little they gain if they do not lose. So I'll just go ahead and publicly put myself in the "mild breeze if any category as well. ^_~

The above hissed in response by: Towering Barbarian [TypeKey Profile Page] at November 2, 2006 11:48 PM

The following hissed in response by: SkyWatch

I Would have put this in a dif topic but all the topics with it are off the radar for now.(mostly the NK ones dealing with Japan)

What makes Japan or anyone else think America can protect Japan now days? We couldn't do crud if Japan got destroyed cept kill those who killed her.We (America) don't have the power to keep it from being destroyed.

Sorry....of topic post and I know you like organization.

The above hissed in response by: SkyWatch [TypeKey Profile Page] at November 3, 2006 12:25 AM

The following hissed in response by: Terrye

sky watch:

We could give Japan the shield.

The above hissed in response by: Terrye [TypeKey Profile Page] at November 3, 2006 2:23 AM

The following hissed in response by: Terrye

I hate to predict, but the polls are so screwey that following them is a lost cause. I think that the GOP will hold the Senate if it loses the House it will be a small margin. I think Sabato is wrong.

The above hissed in response by: Terrye [TypeKey Profile Page] at November 3, 2006 2:24 AM

The following hissed in response by: pete

Hey, a correction about NY 19: John Hall is the democrat and Sue Kelly is the Republican in this race. I know you called the RT strategies/CD poll wacky, this is just one more reason to use this description.

The above hissed in response by: pete [TypeKey Profile Page] at November 3, 2006 6:08 AM

The following hissed in response by: Davod

The Republicans will retain control of the senate and Congress.

The above hissed in response by: Davod [TypeKey Profile Page] at November 3, 2006 6:22 AM

The following hissed in response by: Rovin

Turn-out, turn-out, turn-out! Every republican should already have ONE SIGN posted on their refridgerators. "TAKE TUESDAY OFF, AND VOTE"
AND.....call 25 of 30 of their best (republican)friends and encourage them to go to the polls and vote a straight republican ticket. (enclose nose plugs if needed)

If a strong republican turn-out comes out on Tues, we hold both houses. And the "we wuz robbed" from the moonbats will be the battle-cry.

The above hissed in response by: Rovin [TypeKey Profile Page] at November 3, 2006 7:25 AM

The following hissed in response by: Rovin

Totally-off/on-topic

Ed at CQ has posted a great essay this morning concerning Iraqi Intelligence Service documents and (of course) the NYT's is involved in a way that even the Times admits to authenicate.

The above hissed in response by: Rovin [TypeKey Profile Page] at November 3, 2006 7:41 AM

The following hissed in response by: Dafydd ab Hugh

Pete:

Thanks for the correction; I had just switched the two names. The point itself was correct -- that even the Democratic poll has the Republican ahead, but RT Strategies/CD has the Democrat ahead.

I fixed it.

Dafydd

The above hissed in response by: Dafydd ab Hugh [TypeKey Profile Page] at November 3, 2006 8:23 AM

The following hissed in response by: Big D

Dafydd,

I think you are confusing predictions with polls. I'm sure your predictions (and others) were correct, since you are smart enough to take the B.S. from the polls and apply a little common sense.

So I stand by my comment - the polls are almost always biased toward the Democrats and are almost always wrong. Partisans report these polls as facts, Good pundits, left and right, interpret these polls to reveal the underlying truth.

The above hissed in response by: Big D [TypeKey Profile Page] at November 3, 2006 10:56 AM

The following hissed in response by: Dafydd ab Hugh

Big D:

I think you are confusing predictions with polls.

Another comment; another post!

Dafydd

The above hissed in response by: Dafydd ab Hugh [TypeKey Profile Page] at November 3, 2006 3:32 PM

The following hissed in response by: KarmiCommunist

Honorable High Potentate,

To tell you the truth, humble me wishes that i had never gotten into politics (so late in this ‘Life of Flesh’) in 2002, but i’m stuck now...so to speak whilst planning to Vote this coming Tuesday or at least this last time.

i don’t even know who was favored to win the 2000 Presidential Election, but suspect that Al Gore was...and, do recall that on my drive home from work (whilst listening to the radio) that MSM interrupted what i was listening to, with a prediction that Al Gore had basically won. i live in Florida, don’t work late hours and wasn’t into politics, but thought it a tad strange that Al Gore had won so soon. Frankly, i didn’t care who won back then; however, the Attacks of 911 changed humble me, and i started paying more attention (especially to the rather ‘sTrAnGe’ Modus Operandi way that MSM was reporting the news) to what MSM was saying...or, not saying.

i don’t even recall who was favored in the 2002 Mid-Term Elections, but registered to Vote, and Voted for the Republican Party...in my show of support for President Bush.

BTW, i *DO* recall who was favored to win the 2004 Presidential Election...so to speak of John Kerry.

Humble Low and Ignorant Insane swamp hermit me is probably the most silent of America’s Silent Majority, probably America’s dumbest individual when it comes to understanding polls, but someone besides me Votes against the Democrat Party...so to speak.

MSM continues its Modus Operandi ways...nothing new there. However, recently, John Kerry decided to speak for the Democrat Party once more, and...today, The New York Times showed its desperate support of the Democrat Party.

If the polls are even halfway correct, then how come such October and November ‘slips’ reveal such desperation?!?

Enough said...

The above hissed in response by: KarmiCommunist [TypeKey Profile Page] at November 3, 2006 6:03 PM

Post a comment

Thanks for hissing in, . Now you can slither in with a comment, o wise. (sign out)

(If you haven't hissed a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Hang loose; don't shed your skin!)


Remember me unto the end of days?


© 2005-2009 by Dafydd ab Hugh - All Rights Reserved