State Polls More Favorable For Kerry

One of the current aspects of the media narrative on the race is that while Bush seems to be slightly (2-3%) ahead nationally, Kerry seems to be ahead in battleground states. In the interest of proving nothing but provoking discussion, I spent the last two hours constructing a national poll out of recent state polls. My findings suggest that state polls are indeed more favorable to Kerry than national polls.

The four tracking polls that weight by Party ID to some degree, ABC LV's, Rasmussen, TIPP and Zogby, currently average to Bush 47.6% and Kerry at 45.8%. Using the most recent poll from every state as provided by 2.00k.com, and the 2000 election information provided by Dave Leip, I complied a crude national poll constructed entirely from state polls. After multiplying Bush and Kerry's current standing in a state with the percentage of the national popular vote that state represented in 2000 and then totaling all of the states, Kerry came in with a slight lead of 46.6% to 46.4%.

Seeing Kerry slightly ahead and Bush under 47% brings a smile to my face. If 46.6% to 46.4% were the numbers the night before Election Day, Kerry would win for sure. Of course, many different firms using many different methodologies at many different times conducted the fifty-one polls used in this exercise. This makes my findings pretty salty, but since we are close to midnight, a little salty treat really hits the spot.



Display:


population shifts among states (none / 0)

I love the stats work you do, Chris, and this exercise makes sense except for one bit: population has shifted among states since the 2000 election. Red and pinkish states will have gained the most (NV, AZ, probably NC and GA).  Increasing population in AZ and NV, decreasing population in MI , IA and PA, may account for part of the difference between the average of national RV polls(47 and change to 45 and change) and the 46-and-change tie you computed. (This has nothing to do with the Electoral College math: the point is that the percentage of national popular vote represented by each state will have changed since 2000.) I'm certain the change will be statistically significant in NV, not certain for anywhere else.

The rest of the difference may come from how hard polls push leaners: some state polls seem not to push them much.

by accommodatingly on Mon Oct 18, 2004 at 12:06:21 AM EST

Re: population shifts among states (none / 0)

Yes, but Bush is underperforming in many red states compared to 2000, especially the ones with the highest growth rates (FL, GA and NC). With Kerry up 44-48 in Florida and 51-43 in California, I'm not convined the increases would help Bush at all.
by Chris Bowers on Mon Oct 18, 2004 at 12:44:37 AM EST
[ Parent ]

National v State (none / 0)

For months I've been convinced that this race would turn on the battleground states, not national polls.  Anticipating that Kerry's numbers would be rung up in big states (New York & California) I was worried that what I expected would work to Kerry's disadvantage.  Perhaps it works against Bush.

Your work, Chris, is confirmed in elaborate detail over at 2.00k.com. That site includes a graphic showing the president's chances of garnering 270 electoral college votes.  It is based on a program written by Andrea Amoro (a statistician at the University of Minnesota) Her program checks hourly (!)  for new polls in every state and then runs complex scenarios about thousands of possibilities.  I don't pretend to understand the statistics or the probabilities.

The graphic displays that Bush's chances of cobbling together 270 electoral college votes started sinking through the floor the minute the debates started.  The president had around a 75% chance of re-election in September, a prospect that now stands around 42%.

 

by blaneyboy on Mon Oct 18, 2004 at 02:30:28 AM EST
[ Parent ]

Re: National v State (none / 0)

The url you sited is invalid. Can you supply an accurate url for the research you discuss in this comment?
by leschwartz on Mon Oct 18, 2004 at 05:39:31 AM EST
[ Parent ]

Re: National v State (none / 0)

www.2.004k.com
by Geotpf on Mon Oct 18, 2004 at 06:18:05 AM EST
[ Parent ]

Re: National v State (none / 0)

http://www.econ.umn.edu/~amoro/Research/presprobs.html
by jrrl on Mon Oct 18, 2004 at 10:03:12 AM EST
[ Parent ]

population shifts turn red states purple (none / 0)

Interestingly, the same population migrations responsible for disproportionate growth in many states that Bush won in 2000 is also making these states less reliably Republican.  For example:

  • Mexican, Latin American, and Caribbean (non-Cuban) immigration to the Southwest and Florida.

  • Moderate northern retirees moving to Florida and Arizona.

  • Emigration of liberal Californians to Arizona, Colorado, Nevada, and New Mexico.

Also the Hispanic population, which breaks strongly Democratic outside Florida, has a younger population and higher birthrate than the country as a whole.  This also helps to purple the electorate in the Southwest states.

Bottom line:  Only Texas, Georgia and North Carolina are still reliably Republican among the fastest growing states.

by berith on Mon Oct 18, 2004 at 11:35:17 AM EST
[ Parent ]

Re: population shifts turn red states purple (none / 0)

Actually, both North Carolina and Virginia are populating with "damn yankees" that seem to be strengthening each state's Democratic strongholds (Triangle Area--Raleigh, Durham, Chapel Hill in NC, and Northern Virginia outside DC in VA).  I live in Northern VA, and went to school in NC, so I know these states.  Look for battleground status fop both VA & NC in 2008!  More ups for the Dems on demographics!
by Doug in Virginia on Tue Oct 19, 2004 at 05:43:20 PM EST
[ Parent ]

Not trying to rain on your parade, but... (none / 0)

don't state polls tend to lag behind national trends.
by RandyMI on Mon Oct 18, 2004 at 12:12:51 AM EST

Re: Not trying to rain on your parade, but... (none / 0)

I think that's an urban legend; have read it many times, but not seen it proved. The only lag would be that the polls are later.
by Jerome Armstrong on Mon Oct 18, 2004 at 12:41:54 AM EST
[ Parent ]

Semi Confirmation (none / 0)

Chris, I've been doing much the same calculation using state polls for the last 100 days. I keep a separate calculation for the state polls of Rasmussen, as well as the most recent and next to recent state polls (except throwing away SV and Gallup polls). These have been remarkable more stable than any national poll.

Results are
Rasmussen = Bush + .8%
     This result has never shown the individual state totals to be that close to Rasmussen's daily tracking. I dont think it moved above B+2 or K+2.5 at any time. But it makes me wonder why the discrepency in the Rasy polls.

Latest state polls I factor by the percentage of ev to the total ev and shows Bush +.7.

Next most recent shows K +.99.

Interesting to note however that changes in a state like Texas which can vary by +-10 Bush can change the overall total by .5 to 1 %.  So its a lot more important WHERE the changes are happening.

Currently, I also do battleground calculations (still counting 21 states) which shows B +.6%

I'll try narrowing down to about 10 crucial states to see how it compares!

by citydude2000 on Mon Oct 18, 2004 at 12:45:55 AM EST

Semi Confirmation 12 Battlegrounds (none / 0)

HMMM.

Narrowed Down to 12
ME, MN, IA, PA, WI, FL, OH, AR, NH, NM, CO, NV

Rasmussen K +.3%
Latest State K +1.7%

They're all so close no one state moves the totals very much.

GOTV!

by citydude2000 on Mon Oct 18, 2004 at 01:04:14 AM EST
[ Parent ]

Irreverant View of Today's Senate Races (none / 0)

IRREVERANT VIEW OF TODAYS POLITICAL CLIMATE

Sometimes politics seems so complicated but they it is really very simple.

Florida: Castor-Martinez the two best candidates out of the lots running in a tossup state. So it is then no surprise to say that whomever wins the state up top will likely pull in whomever is below.

Georgia: Majette-very nice lady who should have never given up her House seat. The House could do without McKinney, keep Majette, let Cliff Oxford spend into oblivion trying to make the seat competitive but no one would worry because it was all his money. Meanwhile we have a black fairly liberal candidate in a conservative white Georgia. Unless everyone feels like repenting for past sins its almost a foregone conclusion who will win without even mentioning the other candidate.

Illinois: When the Democratic candidate is getting a third of the conservative vote. Wait I already used too many words on this one...

Oklahoma: Tom Coburn in the white man's version of Alan Keyes. And only in Oklahoma would it be even close. (Maybe in Utah, or Idaho but still...) If Coburn wins it will be because of Oklahoma's innwardness. Only in a place like Oklahoma would rampant lesbianism be taken seriously. Any other state would jump on an attractive candidate like Brad Carson.

South Carolina: Eh I forget about SC in the above statement but to combat rampant lesbianism you need to ban gay teachers so says Jim DeMint. Or let the jobs go overseas because they weren't worthy jobs anyways? And no one minds that a Snickers bar goes from 50 cents to $1.25 because of the national sales tax. But seriously it shouldn't be close but it is because of this SNL fodder.

Alaska: Conservatives love family values, but inbreeding was never widely accepted outside of West Virginia. So it comes as to no surprise that in all of vast Alaska that no one is more qualified than your own daughter. No more breathtaking than the scenery is the fact that this race is even close in Alaska. But as they say "live by daddy die by daddy" With daddy at 37% approval?

Colorado: Getting old is such a bitch. Especially for other people. Because if Sen. Campbell hadn't been feeling age creeping up on him than this wouldn't have been a race. But as it goes there are always death & taxes so suddenly Colorado is in play. The Democrats play their best card in Hispanic 5th generation Ken Salazar. Republicans are left with beer magnate Pete Coors. With a burgeoning Hispanic population, cerveza Jose?

South Dakota: If you fail, try, try again. Animals outnumber people in this state where the population density is measured in the single digits per sq. mile. Like in NH, people here tend to look their candidates in the eye. One-eyed Willie was never popular here but Dakotans sure do love the bacon. Thats why all of them are Democrats. Voters looked Thune in the eye two years ago and didn't like him. So whose to think they will again?

Pennsylvania: Specter represents the closest thing to diversity the Republicans have in the US Senate right now and the closest thing to a Democrat from Pennsylvania. Sure it would be nice to have a D there but you take what you can get. Thats why Toomey lost.

Kentucky: Someone said Washington can make you go nuts. I never believed it until now. Alan Keyes has wacky ideas but he believes in them. So why would a sitting Senator say his opponent looks like one of Saddam's sons and act like no one is paying attention to him? If Bunning loses I will sign myself into a rest home. But at least they saw it coming in IL.

by southerndemnut on Mon Oct 18, 2004 at 02:18:02 AM EST

1980 or 2004 You Decide (none / 0)

1980 or 2004? YOU DECIDE.

Enter an incumbent President who was elected in a very close election over a former VP who was associated with a controversial administration. He was elected largely because he appealed to people's values and that he would reform and clean up the White House and DC. After a somewhat rough start being so new to DC and having to deal with the entrenched power brokers he became very popular after a serious of international and foreign policy coups and for a while was a symbol of American resolve, values, and steadfastness. However as time waned many domestic and international troubles overtook that initial popularity. After a disasterous foreign policy experience with a country named "I" the public was seriously divided over his leadership. Furthermore there were domestic troubles, gas prices were high, the economy had stagnated, deficits were creeping up, and people felt like they weren't going anywhere. Many people felt that our reputation was tarnished around the world. Many people felt it was time for a change. Enter in a one-time darkhorse candidate who many people thought early on wouldn't amount to much, was unable to win, and people just couldn't like. He was scary, he was too serious, he had too many radical idea. He was out of the mainstream. Who would've known?
History tell us what happened in 1980. What will history tell us about 2004?

by southerndemnut on Mon Oct 18, 2004 at 02:19:54 AM EST

Beware of the media narritive (none / 0)

The media is now using the polls which show a slight Bush lead in some national polls to advance the story line that Kerry gained nothing from winning the debates and that Bush in fact is gaining ground on Kerry.

Managing the narrative can make the difference in a close race, and we have seen that surpressing the Democratic vote is very much a part of the GOP and corporate media stratagy.

http://www.cnn.com/2004/ALLPOLITICS/10/17/poll.sunday/index.html

(CNN) -- Although Americans think Sen. John Kerry did the best job in the debates, the Democratic nominee appears to have lost some ground to President Bush in the popularity contest, according to a CNN/USA Today/Gallup poll released Sunday.

The poll interviewed 1,013 adult Americans by telephone Thursday through Saturday, including 942 who identified themselves as registered voters and 788 who indicated they were likely to vote.

In the previous Gallup poll, taken October 9-10 after the second debate on October 8, Kerry and Bush were tied at 48 percent among registered voters. (Full story)

The latest poll, taken after the third and final debate last Wednesday in Tempe, Arizona, indicated an edge of 49 percent to 46 percent for Bush among the same group.

--- end of clip ---

They will broadcast these reults or similar results day after day, every half hour on the headline news service and the other CNN channels until the fabricated narrative becomes a reality in most people minds.

The result will be that some Democrates do become discouraged and make a lesser effort in these next two weeks.

In addition to the next round of the Rove smear campaign these tactics could make the difference in a close race.

by leschwartz on Mon Oct 18, 2004 at 03:20:36 AM EST

Re: Beware of the media narritive (none / 0)

Let's see what the next poll say.  I think these results were influenced by the Cheney remark (bet Kerry doesn't do that again) but that story is being swept off the front page in favor of the draft story (good job John...Smart thinking!!) and the Iraq refusal story.  The tracking polls show a tie, lets see what the next CNN/GOP...I mean Gallup poll shows.
by yitbos96bb on Mon Oct 18, 2004 at 12:31:32 PM EST
[ Parent ]

The anti-coat-tail effect? (none / 0)

Could this discrepancy reflect the changing context of local and national races in a polling script? I.e., if you ask "who do you want for president?" the choices are B v. K. But if you ask, "who do you want controlling Congress and the White House," the choices are Dems vs. Republicans (which means Delay et al as well as Bush).

Could it be that the Republicans look less appealing in this wider context? If this is true, that means that the Kerry campaign gets a bump from local races where Dems are strong, particularly challenger races where there's a swing in momentum towards Dems (as in our PA 6th). Which means that Kerry, with all his wealth of clueless volunteers, should be working really closely with local campaigns, who will know their own ground -- and what sells there -- intimately.

by Anandi in PA on Mon Oct 18, 2004 at 08:47:40 AM EST

Re: The anti-coat-tail effect? (none / 0)

As a former Pennsylvanian transplanted to Northern Virginia, I sure hope you're right about your PA Congressionals, even in the 6th.  Is that Lois?
by Doug in Virginia on Tue Oct 19, 2004 at 05:50:49 PM EST
[ Parent ]

Worked Swing Areas of Iowa this Weekend (none / 0)

We brought a bus load of Kerry volunteers down to Davenport Iowa this weekend to canvass for Kerry. They put us into a "tough" swing area re-IDing independents to see if there was movement. We were closing Kerry supporters 3 to 1 over Bush folks. The debates seemed to be the deciding factor.
by DanielD on Mon Oct 18, 2004 at 07:29:58 PM EST


You are not logged in.

In order to post a comment, you must be logged in. If you have a member account, please log in to comment.

If not, you can make an account right here. It's quick and free.