The latest Zogby poll gives us the bottom line on that info, Jeremy:

Zogby America Poll. Aug. 12-14, 2004. likely voters nationwide


General Election Trial Heats:
Bush/Cheney Kerry/Edwards

ALL 43% 50%

Men 45 47

Women 42 52

But Gallup has it just the opposite.

The Gallup poll was three days earlier :-

Has't Zogby been less reliable than Gallup? The only specific example I remember is the daily polling up to the Clinton-Dole election. He had Dole gaining every day and it looked like it would be close. It wasn't. I haven't paid much attention to Zogby since then.

Zogby today is not Zogby whenever :-) Besides, these are all just polls showing us what people tell the pollsters, not reality.

And actually, the post shows that Kerry is improving his numbers. I blogged on the same topic last April and the Bush numbers show that Kerry is doing 10% better than he was then.

Since I blogged the item, the trade deficit has been getting bigger and bigger, the budget deficit has grown tremendously, and job growth has been shrinking month after month after month.

Kerry looks like a solid winner come November, polls or no polls.

Funny enough, I am polled by Zogby on a regular basis. I can only speak for myself of course, but I do answer the questions as forthrightly as possible. Considering the time required to complete the average Zogby poll, the average participant is probably going to have an interest sufficient to answering the questions honestly.

That all said, the real key to poll accuracy is the questions asked. I have seen polls commissioned for my state, local, and national parties and have seen mistakes made, which give you bad results. For instance, we mistakenly asked questions that did not include all of the candidates on the ballot. That leaves a significant hole in the data, and makes every answer on the question suspect. The biggest flaws I have seen on polls so far have to do with Nader. He is not going to be on the ballot in the majority of states, but will be on the ballot in some. If the poll does not filter for the state of residence, the results will be off. That's just the tip of the iceburg.

What's fun to note is that the supporters of a candidate will show you the poll that has their choice highest. The candidate and the campaign team want the poll that shows him the lowest. That's what you work with.

RE: POLLS

In my past incarnation as a Marketing Director, I was one of the few people in the industry who thought polls were the tools of Satan.

I hated them, but my boss would often want to use these "market surveys" in order to be 'safe' and tell HIS boss that everything pointed in the right direction. (ie, it's a substitute for balls and creativity).

I always bitched and poo-pooed these exercises in wasting time and money, and one day, somebody asked me to defend my criticism.

"Coincidentally," I said, "somebody phoned me last night to ask some 'marketing questions.' I hung up on 'em. So your poll doesn't give you data ON PEOPLE WHO DON'T ANSWER POLLS."

They shut up.

Justene, Shark is right but did you see the latest Gallup poll from Ohio (conducted between August 13-15):

Kerry 50%

Bush 41%

Nader 5%


With Nader not an option:

Kerry 52%

Bush's 42%

I never see polls. (I secretly share Shark's view especially since Zogby told me that Dole would be close). My husband tells me about them. He did start to discuss today's polls when we went to lunch but I cut him off to discuss work, of all things. I'm sure I'll hear soon enough, along with his explanation of why it's not bad. I expect he'll tell me that it is preconvention.