[lbo-talk] Excellent analysis of Time Poll

Doug Henwood dhenwood at panix.com
Sat Sep 4 09:17:12 PDT 2004


Here's a comment from the guy who supervised the poll (with some formatting oddities cleaned up).

Doug

----

From: Mark Schulman Subject: Time Magazine Poll: Additional Comments

I wanted to comment on our Time Magazine Republican Convention Poll, released yesterday. As you know, the poll finds that Bush had an 11 point lead over Kerry, 52-41 among likely voters. We interviewed during the Republican convention, Tuesday thru Thursday, not post-convention.

To quote our favorite line, *polls are a snapshot in time.* We don't know if yesterday's numbers are just the apex of a convention bounce, with the numbers falling back to the ground in a while, or if it marks a real turning point. Previously, the horse-race numbers varied within a fairly tight range, generally in the range of statistical deadlock or very slight Kerry advantage. Yesterday's poll marks the first time in our Time polls that a candidate has broken out of that range.

Please note that our Time Poll the week before the Republican Convention, together with several other polls, including last week's ABC poll, were already finding Kerry slippage, not just in the horserace, but, importantly, in the internal numbers. Among other things, Kerry seemed to go off-message when he got caught up in the Swift Boat accusations and other blistering Republican attacks. His previous edge over Bush on domestic issues, particularly economic stewardship, was already narrowing two weeks ago, for example. And, of course, nominating conventions dominate the news cycle, providing voters with intensive exposure, hence the often-time bounce.

The Time Poll's Bush 'bounce,' comparing last week's numbers to this week's, was a net 9 percentage points among likely voters, and not much different among registered voters. That's a very typical-sized bounce. Please remember that bounces from party conventions are more the rule than the exception. Our Gallup colleagues have put together an excellent history of bounces:

http://www.gallup.com/content/?ci=12919

Convention bounces are no longer assured, as Kerry discovered in early August, getting just a tiny bounce from the Dem convention. The Kerry mini-bounce or no-bounce was a bigger surprise historically than was the Bush bounce that we captured this week.

I'm always pleased to answer any questions about the Time Poll. All of the major election polls are 'open books.' We're all aware of the potential problems lurking in election polling as well, particularly as coverage issues accelerate. We put supporting documents and data on our web site (www.srbi.com) following each poll. There's a slight delay this weekend because of an early start of the holiday weekend at SRBI. I hope to have this week's poll up shortly.

Best wishes, Mark Schulman Schulman, Ronca & Bucuvalas, Inc.



More information about the lbo-talk mailing list