Tuesday: Polls & Predictions

Latest polls, discussion, predictions and predicted MSM spin.

INDIANA

Insider Advantage (May 4)

• Clinton 48%
• Obama 44%

Survey USA (May 2-4)

• Clinton 54%
• Obama 42%

Zogby (May 3-4)

• Obama 44%
• Clinton 42%

DISCUSSION: RCP lists all polls, including the partisan PPP poll and the obscure Suffolk poll, but I stick to name-brand outfits. As you can see, the three most recent mainstream polls for Indiana are pointing in three different directions.

Without knowing any specific of Survey USA's methodology, I suspect their anomalous 12-point edge for Hillary is the result of pushing "leaners." They're the only major polling company that has ever shown a double-digit for Hillary in Indiana. Meanwhile, Zogby's the only major pollster showing Obama ahead in the past two weeks.

Neither of those two outcomes -- a double-digit Hillary win or an Obama victory -- seems likely in Indiana. However, it is an open primary, and Republican crossover votes would break for Hillary. Based on the "Operation Chaos" factor, then . . .

INDIANA PREDICTION

Clinton 54%

Obama 46%

NORTH CAROLINA

SURVEY USA (May 2-4)

• Obama 50%
• Clinton 45%

INSIDER ADVANTAGE (May 4)

• Obama 48%
• Clinton 45%

ZOGBY (May 3-4)

• Obama 48%
• Clinton 40%

DISCUSSION: For several days, people have been asking me if there is a chance Hillary could pull off an upset victory in North Carolina. Never say never, but the chances of such a stunning result are remote.

Look at the RCP poll list and you'll see that only once since February has any poll shown Hillary ahead in Carolina. That was an Insider Advantage poll taken April 29 -- a week after Hillary's Pennsylvania win, and immediately after Jeremiah Wright's media tour, and thus at ebb tide for Obama. Even that poll only showed her ahead 44%-42%, which is to say she had a 2-point margin and was nowhere near 50%. Since then, the same polling company shows Obama's gained 6 points and Hillary's gained 1, meaning the undecideds now seem to be breaking toward Obama.

Unlike Indiana, North Carolina is a closed primary, and the last day to change party registration was April 11, so there's lot less chance of "Operation Chaos" playing a role here.

Now let's talk about what the MSM will obsess over in reporting Tuesday's North Carolina primary results, namely race. According to CNN exit polls, in 2004, 73% of whites in North Carolina voted for Bush; 85% of blacks voted for Kerry. It is estimated that blacks constitute 40% of registered Democrats in North Carolina. And let's face it, if you're a white Democrat in North Carolina -- that is, if you're registered with the Harvey Gantt party -- it's unlikely you've got racial issues.

Based on those factors . . .

NORTH CAROLINA PREDICTION

Obama 55%

Clinton 45%

So, I'm picking Hillary to slightly outperform the polls in Indiana, while Obama significantly outperforms the polls in North Carolina.

Either prediction may be wrong, of course, and I don't have a crystal ball. But if the results are in line with what I'm predicting -- if Obama's margin of victory in N.C. is bigger than Hillary's margin of victory in Indiana -- then you'll see the network commentators talking delegates, delegates, delegates. The reason? North Carolina has 134 delegates and Indiana has 84, therefore a big Obama win in N.C. would have much more impact on the delegate count.

What if Hillary comes on stronger than expected in both states? Suppose that Hillary hits a double-digit win in Indiana and keeps it close in North Carolina. Then the MSM talking heads will be all about "electability": Does this result show that Obama was critically damaged by the Wright controversy?

Now, what if Obama exceeds expectations in both states? If he keeps it close in Indiana and really scores a blowout in Carolina -- getting 56% or more -- then it's going to be Hillary Death Watch on the networks. Those pundits will be driving nails in her coffin so fast it'll sound like a hammering contest.

I'll be live-blogging Tuesday night's results, perhaps moving Crisis '08 headquarters a remote location to be divulged later. So be sure to come back and laugh at me when my predictions turn out to be completely wrong, as they almost invariably are.