vt.Buzz ~ a political blog

Political notes from Free Press staff writers Terri Hallenbeck, Sam Hemingway and Nancy Remsen

9.26.2006

New numbers

If a new Rasmussen Reports poll is to be believed, Republican Senate candidate Rich Tarrant might want to rethink the direction of his full-court-press television ad campaign against independent Bernie Sanders.

According to the latest Rasmussen poll, Sanders has actually increased an already large lead over Tarrant by four more percentage points since Aug. 3, the last time Rasmussen polled Vermont.

The new poll said 64 percent of the 500 likely voters surveyed would vote for Sanders and 32 percent would support Tarrant. In August, Sanders had a 28 percentage point leade. In late June, Sanders Rasmussen had Sanders ahead by 38 percentage points.

Rasmussen also found that Democrat Scudder Parker had pulled closer to incumbent Republican Jim Douglas in the gubernatorial race and that Democrat Peter Welch was leading Republican Martha Rainville by 8 percentage points in the open-seat House race. The poll was conducted Sunday and has margin of error of 4.5 percentage points

Douglas still had a double-digit lead over Parker, with 52 percent to Parker's 38 percent. But Douglas' 14 percentage in the latest Rasmussen poll was down from the 18 percentage point he had in August and 23 percentage point lead he enjoyed in the June Rasmussen survey.

This was the first time Rasmussen had polled the House race. Welch had 51 percent support of those polled and Rainville had 43 percent support.

Anyone want to make some sense of all this?

-- Sam Hemingway

posted by vt.Buzz | 1:05 PM |

Comments:

This Rasmussen poll is more evidence that the ARG poll was bogus. The ARG poll showed Tarrant behing by only 15 points.

It's time that American Reserch Group was exposed. It is a New Hampshire based polling organization with Republican ties. It produces polls that are supposed to be independent, but are actually used to mislead people.

# Posted by:

Anonymous | 1:23 PM, September 26, 2006

So the Scudder / Douglas race is closer than the Tarrant / Sanders race and yet the media concentrate on the senate race?

Why?

# Posted by:

Anonymous | 1:43 PM, September 26, 2006

Three different polls, all different. One of them is bound to be closer to the truth. Bernie's people will say the second. Rich's people will say the first. Bernie's will say this one is pretty accuratly showing what there polls say.

I hate polls. Too much room for error. I don't think they matter anyways until you get two weeks out or so.

We'll see, the only polls that REALLY matter are on election day.

# Posted by:

Anonymous | 1:50 PM, September 26, 2006

There's nothing misleading about the ARG poll: Support for Bernie had not eroded at all.

Scudder needs to get out on the TV.

# Posted by:

Anonymous | 1:52 PM, September 26, 2006

Sorry Simms, but I don't buy your explanation of differing polls with differing results.

The ARG polled showed up at a very convenient moment for Tarrant. ARG was unclear about who paid for the poll. ARG's polling numbers are dramatically different from all other polls.

Doesn't pass the smell test.

# Posted by:

Anonymous | 1:54 PM, September 26, 2006

You don't have to buy my explination. I'm not trying to sell it, just carrying on a conversation.

You can't really compare any poll with another. Whether numbers are higher or lower then another, really doesn't matter.

You have to compare the ARG poll with the first ARG poll, and the first Resmussen Poll with teh Second.

You can't compare the ARG poll to WCAX. Or WCAX to Rasmussen. There's too much of a difference, different questions, different samples. etc.

# Posted by:

Anonymous | 3:47 PM, September 26, 2006

Raum Emanuel is as pitiful as Welch is.

# Posted by:

Anonymous | 4:51 PM, September 26, 2006

Anyone marginally familiar with political polling knows that weekend polls ALWAYS overpoll in favor of Dems. This Rasmussen poll was a Sunday poll. I hope you Dems sit back and bask in the glory of that poll, a la Doug Racine.

# Posted by:

BarreBoy | 5:46 PM, September 26, 2006

Sure, Sam, I'd love to make sense of those numbers, right after you make sense of these other numbers for the Sanders/Tarrant race:

Tarrant / Sanders / Firm / Date / Margin
40 / 55 / ARG / 9/13-9/14 / 4.0
27 / 66 / Garin-Hart-Yang (D) / 9/15-9/17 / 4.4
33 / 58 / Research 2000 / 9/18-9/19 / 4.0

All taken within 6 days of each other, one survey has Sanders up by 15, the next up 39, and the last up by 25. How does that make sense?

The fact is that polls done on the cheap -- like Rasmussen, Research 2000 and ARG -- are wildly erratic and their results are famously hit or miss. My bet: put your money on Sanders internal polling (Garin-Hart-Yang), which also shows Welch: 47, Rainville: 41 and Douglas: 54, Parker 31.

General rule: don't get too excited about polls that blast out the results online. With polling, you get what you pay for.

Another good rule: before you write a story for the Free Press on this poll, talk to two professional pollsters, one at a Dem firm and one at a Rep firm. Or better yet, have the Free Press pay for a real poll by a reputable political polling firm.

If my memory serves, Free Press reporting on a poll for the Governor's race four years was way, way, way off. Perhaps you could speak with the reporter on that story and see if there are any lessons learned.

Just a thought.

# Posted by:

Anonymous | 6:11 PM, September 26, 2006

Of these four polls, the ARG poll is the only one that is really different:

Tarrant / Sanders / Firm / Date / Margin

40 / 55 / ARG / Sept 13,14 / 4.0
27 / 66 / Garin-Hart-Yang (D) / Sept 15,17 / 4.4
33 / 58 / Research 2000 / Sept 18,19 / 4.0
32 / 64 / Rasmussen / Sept 24 / 4.5

Other than ARG, the rest are all pretty close to each other, within their margin of error.

More evidence that ARG poll was a bogus poll probably somehow supported by the Tarrant camp.

# Posted by:

Anonymous | 6:56 PM, September 26, 2006

ARG's results were closer to R2K's than Bernie's. I guess that "proves" that Bernie's poll was a fraud.

# Posted by:

Anonymous | 7:14 PM, September 26, 2006

Every poll shows Bernie with a double digit lead.

So who gives a rat sass?

# Posted by:

Anonymous | 7:57 PM, September 26, 2006

The governor's race seems to be getting closer as the debates are getting underway and Parker's getting his message on the airwaves, but I bet the full text of that poll shows still not an overwhelming number of Vermonters know the name "Scudder" (seriously, pick up a newspaper.)

# Posted by:

Anonymous | 8:31 PM, September 26, 2006

BB Wonder,

The Welch-Rainville is certainly close and should be getting much more media attention. But if you look at the so-called right leaning poll ARG, it shows that Welch should have much more room to widen his lead since he is only getting 66% of Dems. Welch is in the lead and it should grow if he can keep his lead with Indies. I think the new ads that are more about what Welch stands for rather than just I'm not a Republican are made for Indies.

# Posted by:

El Guapo's Ghost | 9:21 PM, September 26, 2006