A Weblog of Centrist Voices in American Politics
|
|
|
A Weblog of Centrist Voices in American Politics | |
|
July 17, 2004Poll Watching--"the Bounce"The usual disclaimer: Polls are not always accurate, and this far out they don't really mean all that much before both conventions have been held, and they're pretty variable even in the final run-up. The only poll that really counts is the one taken at the ballot boxes. The "popular vote" doesn't mean a damn thing in Presidential elections--all that counts is the Electoral College vote, as Al Gore can tell you. "The Bounce" has finally materialized now that the polls have had a chance to catch up to the news. This has resulted in some of the swing states moving over to Kerry from Bush. My current EC vote projection using current polls is Kerry 311, Bush 216, with Missouri too close to call. Removing states where the lead is within the margin of error, the numbers are Kerry 249, Bush 205. Currently "in play" by margin of error are: Minnesota, Iowa, Missouri, Nevada, Ohio, Florida, and Maine. Kerry's numbers solidified in a few states, but declined in other swing states, most notably in Iowa, Minnesota, Maine, and Florida. Bush's numbers improved marginally in Nevada, Arkansas, and West Virginia. Edward's home state of North Carolina remains Bush's outside the margin of error, but did tighten up after the Veep announcement. As always, these numbers WILL shift throughout the election season. Stay tuned! Posted by Tully at July 17, 2004 10:57 AMComments
I found this interesting (PDF!): http://www.fabmac.com/FMA-2004-07-14-Approval-Gap.pdf These GOP consultants have identified the problem and have a solution: 1) Drive up Bush's approval number and 2) help voters see Kerry in a negative light. At this point in an election year, #1 is more difficult than #2. Plus it's safe to assume Karl Rove will make the decision, thus making it a very safe bet that tons of mud will be thrown to see how much sticks. Problem for Bush/Cheney is that, if we know anything, it's that Kerry/Edwards will pick it up and throw it right back. Carry an umbrella, folks !!! Posted by: Erasmus at July 17, 2004 11:39 AMThe subject of this thread is poll-watching. It's a specific topic with many broad approaches for discussion. Exactly ONE of the above three comments is somewhat on topic. (Thanks, Erasmus! Interesting link.) The other two comments are being deleted for reasons of thread-hijacking, having absolutely nothing at all to do with the thread subject. They are being treated as comment-spam. If you want to leave the subject entirely to pimp a third-party candidate or spark an off-topic debate on your preconceived notions, do it elsewhere. Posted by: Tully at July 17, 2004 05:07 PMUmm...at the risk of incurring the wrath of Tully... Go Kerry/Edwards! : ) By the way...we just got an inside scoop from a very reliable DC source....check out PK. Posted by: Carla at July 17, 2004 08:07 PMWell, Carla, at least you referred us to a polling story on your own blog! But you just gotta do those link inserts for folk who don't know the way.... PK = Preemptive Karma Blog (We can do a candidate cheering thread later, when the vote is upon us.) Posted by: Tully at July 17, 2004 08:23 PMTully: Okie dokie. I'll do that next time. When I posted our blog link on some other blog comments some of the people got a bit snitty. I never seem to know when it's okay or not. LOL I think I need the "Ms. Manners Book of Netiquette". Posted by: carla at July 17, 2004 10:37 PMFellas: I referenced this post in a piece on PreemptiveKarma: http://preemptivekarma.com/2004/07/whos-right-whos-wrong.html I tried to do a trackback for it...but apparently my less than stellar techy skills are keeping it from happening..? Anyway...good intentions. Posted by: carla at July 18, 2004 01:59 AMIt's funny how even the best poll watchers, those who understand their many limits, like to make interpretations. (BTW, Not directed at you Tully.) One aspect of margin of error in polling that's often missed or overlooked is that a change from week to week (poll to poll) that's within the margin of error might not really even be a change at all : 47 +/-3 means "around 44 to 50, " and 49+/-3 means "around "46 to 52." What this means is that any one time variance in numbers shouldn't be regarded as a change per se. Unless the change is greater than the MOE, the variance could be random static.When multiple polls show similar results, or a 3 or 4 week series shows a trend, that's when you can say "hmm, something might be going on." One thing Tully hints at that hasn't been dealt with much is the idea that poll results tend to lag the news. For example, some people were looking for Kerry to get an Edwards bounce based on polls done just about right after the announcement, and that's too soon. And for things like bad Iraq news or good econ news, this stuff tends to trickle through. IMO, you8 need to factor in a lag of a week or two at this point in time. Maybe later as the focus on the election becomes greater you'll see less lag, but it's sensible not to expect polls to always show immediate response to news events. Posted by: bk at July 19, 2004 09:19 AMI'm hoping Bush wins the popular vote and Kerry that electoral vote, just to drive everyone nuts. Posted by: Oberon at July 19, 2004 11:04 AMI'm with you Oberon, That way everyone who left a paper trail could be hoist by their own petards. We could see record numbers of hoistings. Posted by: bk at July 19, 2004 12:32 PMI'm hoping Bush wins the popular vote and Kerry that electoral vote, just to drive everyone nuts. And Kerry needs to win a swing state by around 500 votes, just to make election night extra fun. It was Florida's turn last time -- so why don't we do it right here where I live in Ohio this time? Posted by: William Swann at July 19, 2004 01:47 PMBryan's spot on about the poll lags and MOE's. Especially when reveiwing state polls, whch tend to come out sporadically and weeks apart, it's very important to watch the timelines. And as I keep telling local politicians, what a poll says is that the people polled, if they were telling the pollsters the truth, would have voted the way they stated plus/minus the MOE at that point in time, if they actually voted, with a 95% confidence level. There's still that 5%, and there's still that margin of error, and there's still sampling error, and.... You hope you have a representative sample, and that your poll isn't an "outlier." You hope you polled actual voters, not just "registered couch potatoes." And you hope they didn't lie to you, or answer whimsically, or on a flip-a-coin whim. Properly done, statistics don't lie. But they only speak very narrowly, and are easy to misinterpret. Posted by: Tully at July 20, 2004 11:02 PM |
Archives
March 2006
February 2006 January 2006 December 2005 November 2005 October 2005 September 2005 August 2005 July 2005 June 2005 May 2005 April 2005 March 2005 February 2005 January 2005 December 2004 November 2004 October 2004 September 2004 August 2004 July 2004 June 2004 May 2004 April 2004 March 2004 February 2004 January 2004 December 2003 November 2003 October 2003 September 2003 August 2003 July 2003 June 2003 May 2003 April 2003
Recent Entries
Dubai Out
Why So Long Between Democracies? Round One, Centrism Rock Lobster? Blackwell Releases "Worst-Treated" List "IRV" used in Burl., VT for mayor election. Great idea! Random Thread Election 2006: Round One A Proper Multiculturalism Bush proposes line item veto act - what's changed?
|
