Pollster.com

Articles and Analysis

 

"Outliers" for 3-19

Topics: 2008 , Frank Newport , Gary Langer , Hillary Clinton , Iraq , John McCain , Mark Mellman , Pollster

Poblano puts up a new site (538) that aggregates polls weighted with a "reliability rating."

Gregory Nini and Glenn Hurwitz produce a statistical model and conclude that "roughly two million more people would have voted in Florida and Michigan had they expected their delegates to be seated" (via Smith, Swampland).

Geoff Garin gets added to the Pollster roster of the Clinton campaign.

Mark Mellman is starting to worry about a divided Democratic Party.

Josh Goodman thinks Democrats should use a poll to allot delegates in Florida and, somewhat surprisingly...

David Hill floats a similar idea (though advises against calling it a "poll").

Carl Bialik sees problems with the Clinton big state electoral vote theory.

Gary Langer looks at the changes in attitudes on the war in Iraq over the last five years.

Jennifer Agiesta posts a similar retrospective.

Frank Newport shows that Americans disagree that going to war with Iraq was the "right decision"

Peter Feaver advises John McCain to defend that decision (via Ambinder, some background here).

 

Comments
Joseph E:

Everyon should check out fivethirtyeight.com
Poblano has put together some amazing graphics and maps.

Right now, he has Obama with about a 55% chance to win, and Clinton with a 40% win percentage against McCain, based on state-level polls re: the general election. I can't vouch for his methodology 100%, but it seems the best thing out there. Certainly much better than SurveyUSA's misleading maps, which gave both Democrats too much credit for close races.

It will become a very useful resource when the primary campaign wraps up.

____________________

Carl Bialik made one major mistake in his analysis:

Hillary Clinton has NOT won 8 of the 9 largest states.

The ninth largest state is Georgia, which Obama won handily. New Jersey is #11.

If you leave out Florida and Michigan (which don't count, especially in light of the excellent analysis of Nini and Hurwitz) and generously count Texas as a tie, you have among the top nine a net record of 3-2-1 so far.

I realize that Pennsylvania is polling for Hillary, but we shouldn't forget that if you ditch the arbitrary "Top Nine" nonsense, you'd see that North Carolina is #10 - and it's polling heavily for Obama.

Not only is the big state argument utterly wrong from the start, but it is constantly being repeated by reporters like Bialik without the smallest amount of fact-checking. His article even gave a link to the data that proves the analysis was fundamentally flawed!

This is ridiculous. These reporters should be ashamed - if not fired.

____________________

Gary Kilbride:

Regarding Mark Mellman's piece, he shouldn't worry. Everything does tend to drift back to the beginning with big picture overcoming family feuds, and this is a Democratic year. Ask any GOP operative, to use a Bob Novak term, if they wouldn't gleefully trade playing fields. Only the ones who aren't qualified to be there in the first place would say no.

Look at the years Mellman listed, the ones with a so-called divided primary. In every instance other than 1980, the situational influence naturally favored the other side to begin with.

* 1964 (GOP divided) -- Democratic advantage, one term in office and very popular president recently slain

* 1968 (Democrats divided) -- Republican advantage, two straight terms for Democrats and unpopular war

* 1972 (Democrats divided) -- Republican advantage, incumbent with party in power only one term, the most favorable scenario imaginable

* 1976 (Republicans divided) -- Democratic advantage, two straight terms for Republicans, and Watergate

1980 (Democrats divided) -- Democratic advantage, and the one glaring exception, incumbent with party in power only one term, this the only loss in last 10 examples, demonstrates a horrid economy and overseas turmoil can negate even the most dramatic situational edge

1984 (Democrats divided) -- Republican advantage, incumbent in place and party in power only one term

Plus, Mellman properly identified 1952, with a divided GOP but two decades of continuous Democratic rule in the White House. That had to be a Republican edge, even without a war hero as nominee.

And 2000 was always a GOP edge, no matter how Democrats thrilled to deny it. Republican incumbent, party in power one term, war time, approval ratings hardly abysmal, economic numbers solid in the traditionally key areas. I'll always maintain the Swift Boat episode was the most laughably overstated "factor" in modern election history. Cite any poll you want, scrambling to rationalize. Irrelevant. Kerry was destined to lose a close race regardless. Swift Boat is a convenient reference point for lazy analysts, along with unwillingness to credit Bush for finally winning legit.

Fast forward to '08, two straight terms for the GOP, a president who can't bust mid 30s in approval rating, and debate whether the proper term is recession or depression. Who are we trying to kid? Anything is losable in a relentlessly partisan era, but I'll play that hand every time.


____________________

Gary Kilbride:

Whoops, I obviously meant 2004, not 2000, when discussing Swift Boating.

But 2000 is interesting because the advantage was uncertain. Two straight terms Democratic terms favored the GOP but economic conditions and overall contentment were advantage to Democrats.

Looking to 2012, it should be a Democratic tilt regardless of November's result. You'll either have a Democratic incumbent with his/her party in power only one term, or the GOP struggling to hang on after three consecutive terms in power, the first time we've had that itchy scenario since Bush 41 was ousted by Clinton in '92.

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR