Tuesday, September 21, 2004

Battleground State Polls

There's been quite a bit of blasting back and forth about how the race is shaping up in the battleground states. My sense is that the race is very fluid. Some recent polls have Bush leading in several of the battlegrounds that Gore won in 2000--IA, MN, WI, etc. On the other hand, ARG just came out with the first 20 of their 50 state polls and Kerry seems to be doing quite well:

CO: Bush 46, Kerry 45
ME: Bush 44, Kerry 48
MN: Bush 45, Kerry 47
OR: Bush 45, Kerry 47
WA: Bush 44, Kerry 51

This isn't to say that I'm shilling for Kerry. It's just that you need to take these state polls with a grain of salt at this point. First, many of them are days and weeks old, and don't reflect the current situation. Second, many of them have relatively small sample sizes, making the margin of error quite large. For example, a sample of 600 people (as in the ARG polls above), gives a MOE of +/- 4 percentage points 95 percent of the time. This means that any of the numbers listed above for Bush or Kerry could be 4 points larger or smaller. Take CO, where Bush has 46 and Kerry 45. Since the MOE is +/- 4, Bush could be leading handily by 50 to 41, or Kerry could be ahead 49-42. And that's 95 percent of the time. The other 5 percent of the time, Bush or Kerry could be ahead or behind by even more.

All in all, I think Bush probably has a bit of national lead, somewhere around 5 points. Even with a relatively small lead like that, a number of the battlegrounds, including several of those won by Gore in 2000, will start to move his way. If Kerry closes the gap, those states will swing back to him. So while it's clearly better to be up rather than down, this isn't a enough of a lead that Bush can start to coast. At this point in 2000, Gore was leading Bush by about the same amount. I was also struck by the finding of one recent poll (Pew, I think) that the number of undecided voters has gone up in recent weeks, when it usually goes down once the campaigns really start. So, can Kerry catch up? Sure. Will he catch up? Damned if I know. Ask me on November 3.

One other point. One of our regular commenters keeps wanting to blast the Zogby polls since they seem to have Kerry doing well in several battleground states. But Zogby's results have also been born out by other polls (see above). Plus, I remember how in 1996 Zogby got all sorts of grief from Democrats because he consistently showed Dole doing better than in other polls. In the end, it turned out that he's was right. Not only did he beat the other pollsters in predicting the final outcome of the presidential race, but he nailed the final numbers exactly on the head. Interesting now that his polls seem to be showing Kerry running better than some other polls, he's now getting it from Republicans. (Full disclosure: John Zogby is based in Utica, NY, where I live, and I know him pretty well. Over the past few years I've done some work for him and also contracted to do some polls with his firm for the public affairs center at Hamilton. John's a smart, professional guy and his firm does good work. If his polls turn out to be wrong, it will be for all the usual sorts of random reasons. The thought that he's putting his thumb on the scales for Kerry is ridiculous.)


Palooka said...

Zogby is definitely an outlier. His polls are between 5 and 8 points off the average.

Gallup's reported 13 point lead is as much of an outlier as Zogby, but I don't see you reporting that number. I don't see you defenind that number.

Gallup was within two in 2000. And Zogby was within 2 in 2000. Zogby overstated Gore's lead, Gallup understated. But both were, more or less, fairly accurate.

Oh, if within four points is supposed to mean it's compeitive, why haven't you reported the Gallup poll showing the race within four points in NY?

Let me guess, Zogby's been whispering sweet nothings into your ear.

Please, continue building your Zogby-equipped cocoon. I have no problem with that. Maybe ya'll will believe it's "Kerry's race to lose again."

Anonymous said...

Palooka is a tool.

Anonymous said...

I'm not someone who's blasted Zogby's polls, but it should be noted that Zogby's state polls all use his new internet methodology (the "Zogby Interactive" label identifies these). Whether that methodology is as reliable as the methodology used by other polling organizations is, it seems to me, an open question.

What does that mean? That I couldn't justify relying on his internet polls when those polls show results at odds with telephone polling, whether I like the results or not.

Anonymous said...

Gallup or Zogby, Gallup or Zogby...Lions and Tigers and Bears, Oh My!

Seriously, pick your poison. More than one and probably all of these polls are wrong (and by wrong I mean off by more than the margin of error). But it doesn't require much more than opposable thumbs to KNOW that NY is not going in the Red State column.

Marketing man said...

You have a great blog here! I will be sure to book mark you.
I have a cost in lead lead leads.com low opt phone real time verified site. It pretty much covers cost in lead lead leads.com low opt phone real time verified related stuff. Check it out if you get time :-)