On February 13th, Rich Miller published a full run-down on the Survey USA in the Capitol Fax (sorry, but you have to subscribe to the actual fax for this one–see site for how).
The gist of the story is that not only is Blair Hull taking a lead, but he is doing it in almost every category. Take a look at the results available from Survey USA (pdf format). Hull isn’t just ahead overall, he is ahead in all of the demographic categories except Black and Hispanic. In those two, he is second. Geographically, Hull is ahead in Downstate and the Collar Counties, second in Cook behind Obama. Amongst those ideologically self-identified, he’s tied Obama amongst liberals and ahead with conservatives and moderates. As Miller pointed out, Hull is liberal, but no one ever accused mass public opinion of being consistent.
A small note on the Survey USA polling–it is done by computer so the reliability is questioned by some.
Many seem surprised by the findings, but why? There is a model for his campaign (other than the statistical one he devised) and it was John Corzine’s campaign in the 2000 New Jersey Senate Election.
Corzine faced former Democratic New Jersey Governor Jim Florio in the primary. Early on, Florio was well ahead with both organizational support (though key players went with Corzine) and name recognition. Corzine won 58-42.
Let’s look back at the polls in that race.
February 23, 2000 35 point lead for Florio according to a Quinnipiac Survey. That was about 3 1/2 months out of the June 6th primary election with articles noting that Corzine had spent over $3 million
March 30th 50-26 Florio lead
May 10th-Just under one month before the election 48-33 Corzine over Florio
May 26th 47-28 Corzine over Florio, 56-30 for Corzine of those who intend to vote
Final tally 58-42 Corzine.
For one obvious reason this race is different. There are five serious contenders in the race instead of two, but let’s look at the polling trends since September 2003.
September 2003–5 1/2 months from election day
Hynes 10
Hull 9
Obama 8
Chico 7
Pappas not included
January 2004 2 months out
Hynes 14
Obama 14
Pappas 14
Hull 10
Survey USA January
Hynes 20
Obama 19
Hull 19
Pappas 18
Survey USA February
Hull 29
Hynes 19
Obama 19
Pappas 14
One month out Corzine showed distance. One month out, Hull is showing distance.
Does this mean Hull will win the primary? No. The field is divided so Hull has a more dynamic race to run than Corzine did. It does mean that someone who is getting hurt by him has to find a way to attack him. And that person is Dan Hynes, as Miller pointed out in his Friday story. Clearly, Hynes’ internals were showing deep problems downstate and hence the recent tour of Central and Southern Illinois. Free media can be picked up in smaller markets–even a little time in Metro-East.
The larger problem is that Hull hasn’t even hit his big media buy yet and Dan Hynes is in no shape to counter that effort. He doesn’t have adequate funds. Negative ads are a problem because Hynes would have to take ownership of the ads according to the new FEC regulations and say he paid for the ad (Miller pointed this out too). And Hynes agreed to Dick Durbin’s no negative ads against other Democrats pledge.
In fact, if Hynes were to attack he would probably damage Hull to some degree to the benefit of Barack Obama who can stay above the fray and simply look to turnout black and liberal votes while the other two go after one another. In such a scenario, Joyce Washington may play a spoiler to Obama even with a small number of votes.
Why did so many miss this? Because the past guides most of our views of the future and in the past, organization wins in Illinois. And Dan Hynes has organization as my e-mail demonstrates with the union of the day endorsement for Hynes. When one looks at other states, the potential for Hull was clear. New Jersey is heavy with organization, but Corzine was able to create his own organization with money–just as Hull is doing.
Fortunately, I haven’t been a naysayer. Since I’ve been doing the Roll Call in June, I’ve never had Hull above 2 or below 3. Why? I pointed out organization wins in Illinois, but money can buy an organization.
Unfortunately, a grave bias against stats geeks still exists in the political world,
From the Sun-Times on January 22nd, BLAIR HULL’S WINNING FORMULA
The densely written theorem is a string of precise numbers carried out to the eighth decimal place. It’s designed to identify people for Hull to target with fliers in the mail or telephone calls. Those chores were previously handled by precinct captains, who often flunked high school math.
"The basic idea," Hull said, "is you take the probability of voting times the probability of voting for Blair Hull and multiply them together, and that gives you the relative weight of how desirable it is to contact that person."
For those paying attention to modern campaigns, this is nothing earth shattering except that it formalizes the system of identifying certain and likely supporters in a formula.
This is especially useful if one is trying to create a get out the vote (GOTV) organization out of nothing because one can target 1s and 2s, as certain and likely voters are called, in the most efficient manner possible. Some folks didn’t pick up on the point of the formula in the article:
"The formula says ‘I’m spending $40 million,’ " Pappas said, laughing uproariously after learning of his formula. "He’s driving everybody crazy in this race, except me.
"Everybody will be running around trying to figure out his formula. … Forget about the formula. The formula is: The guy’s got money."
Not only that, but he is trying to use it efficiently.
The most professionally embarrassing comment comes here:
Michael Mezey, a political science professor at DePaul University, read the Atlantic Monthly piece but is skeptical.
"Winning primaries is very much sort of getting your people to the polls, especially in a divided race," he said. "And that’s not subject to a mathematical formula."
Some of the old guys just don’t get what statistics can do. The whole point of the model is to focus efforts on the most likely voters for Hull which also includes the GOTV efforts. Given Hull is already canvassing downstate, I’d say he’s planned for this.
This doesn’t mean a formula alone will do it, but it can be a powerful tool in campaign management. I cannot evaluate it without the data and model and some free time I don’t have, but there is no reason to dismiss it.
Later today, will be a new Roll Call (hopefully). I think you can guess who is on top.