In 2000, Bush lost the popular vote by about half a million votes. In 2004, Bush won the popular vote by 3.5 million votes. That’s a shift in Bush’s direction of 4 million net votes.
Where did this shift in margin–these 4 million votes–come from?
It is possible to answer this question by comparing Bush’s margin in individual states in 2000 with his margins in those same states in this election. This analysis shows the following:
1. About half of Bush’s gains came from the solid red states–those states that gave Bush a margin of 6 or more points in 2000. And about half of these gains in the solid red states (a quarter of Bush’s total gains) came in just four specific states: Texas, Oklahoma, Alabama and Georgia.
2. About a third of Bush’s gains came from the solid blue states–those states Gore carried by 6 points or more in 2000. (In these states, Bush gained by reducing his deficits relative to 2000). And about three-quarters of Bush’s gains in these solid blue states came from just three states: New York, New Jersey and California.
3. About a fifth of Bush’s gains came from the “purple states”–those states that were decided in 2000 by less than 6 points (which includes almost all of the 2004 swing states). And almost all of Bush’s gains in this group of states come from just two states: Florida and Tennessee.
Coming soon: analysis of the county-level vote.
TDS Strategy Memos
Latest Research from:

Editor’s Corner
By Ed Kilgore
-
January 9: California’s Crowded Gubernatorial Race a Bit Perilous for Democrats
As a registered voter in California, I’ve been watching the slowly developing 2026 gubernatorial race in which no Democrat seems to be breaking out of the bipartisan pack. I wrote an early assessment for New York:
The last three governors of California were all legendary, larger-than-life political figures. Arnold Schwarzenegger (2003–’11) was a huge Hollywood and pop-culture celebrity before he entered politics in a recall election that ejected his predecessor Gray Davis. He remains the last Republican to be elected as governor or U.S. senator in the Golden State. Jerry Brown (2011-2019) served in his second two-term gubernatorial stretch, having first been elected to the office way back in 1974 (he also ran for president three times). And the current and outgoing California governor, Gavin Newsom (2019-present), was San Francisco mayor and two-term lieutenant governor before stepping up to the top job in Sacramento. He, too, has dominated California politics in a big way.
The contest to choose the 41st governor of California currently has ten candidates — eight Democrats and two Republicans — and not that many voters could identify them in a line-up. Two Democratic politicians who did have some name ID and who might have dominated the field have given the race a pass. That would be former U.S. senator, vice president, and presidential nominee Kamala Harris, who may instead run for president again in 2028 (very likely against Newsom); and her successor in the Senate, Alex Padilla, who gained a lot of attention when he was wrestled to the ground and handcuffed by Secret Service agents for trying to ask Homeland Security Secretary Kristi Noem a question at a press conference.
With those big fish out of the tank, the remaining field is composed of candidates who are far from unknown, but are still small fry, relatively speaking. A well-known former Democratic member of the U.S. House, Katie Porter (who ran for the Senate in 2024) and current House member Eric Swalwell (who very briefly ran for president in 2020), are running. One current Democratic statewide office-holder, Superintendent of Public Instruction Tony Thurmond, is making a bid. So is former state comptroller Betty Yee, former Los Angeles Mayor Antonio Villaraigosa, former Biden administration HHS Secretary Xavier Becerra, and former state assembly majority leader Ian Calderon. The most recent Democrat to enter the race was hedge fund billionaire and liberal activist Tom Steyer (who ran a presidential campaign briefly more successful than Swalwell’s in 2020).
Alongside these eight Democrats are two Republicans: Fox News gabber and former British Tory political operative Steve Hilton, and current Riverside County (east of L.A.) sheriff Chad Bianco.
Polls consistently show these ten candidates struggling to break out of the pack. Early on, Porter, building on name ID from her unsuccessful 2024 Senate race, had some buzz, but she damaged herself by pitching a temper tantrum during a media interview that wasn’t going her way. Since then it’s become a sluggish race between snails. The latest public poll, from Emerson, released in early December, shows Bianco at 13 percent, Hilton and Swalwell at 12 percent, Porter at 11 percent, Villaraigosa at 5 percent, and Steyer and Becerra at 4 percent. The remaining candidates combine for 7 percent, and there’s an impressive 31 percent who are undecided or don’t know who these people are. Everyone but Porter has name ID under 50 percent, and hers isn’t all that positive. You may think that’s because it’s so very early in the contest, but in fact, the primary is on June 2, just over six months away.
That primary, by the way, is part of California’s non-partisan top two system in which the first- and second-place finishers, regardless of party, proceed to the general election. And the early polling has created a bit of a freak-out among Democrats bewailing their candidates’ lack of star power, as Politico noted:
“California Democrats have a math problem: They’ve added so many candidates in the race to succeed Gavin Newsom that two Republicans could end up winning the state’s quirky ‘jungle primary,’ shutting the Democrats out.
“A Democratic wipeout is still unlikely. But the prospect of a humiliating pile-up, with no clear powerbroker to act as traffic cop, has put the state’s political class increasingly on edge with each new entrant into the field.”
Even though the race should intensify considerably as we get deeper into 2026, the candidate filing deadline isn’t until March. So the power vacuum in the gubernatorial field could yet attract a late entry from some celebrity (Hollywood is chock full of them) or insanely rich self-funder (one such bag of money, Los Angeles developer Rick Caruso, could run for governor if he doesn’t run again for L.A. mayor). Or more Lilliputs could join the race hoping that lightning strikes (e.g., state Attorney General Rob Bonta).
If the field remains as it is, keep an eye on Steyer, whose vast wealth could buy him the name ID he needs. Ideological divisions and factional alignments could also be key. Thurmond is touting his support for a single-payer health care system and has the endorsement by California’s powerful teachers unions. Villaraigosa (who ran unsuccessfully for governor in 2018) has a well-worn reputation as a Democratic “moderate.” Porter has scars from her battles with the crypto industry, which savaged her with negative ads in 2024, while Calderon has become a crypto bro ally. Becerra can run on his legal battles with the first Trump administration (when he served as California attorney general) and Swalwell has been trading insults with Trump for years. Meanwhile the two Republicans in the race can be expected to compete for a Trump endorsement (Hilton is a long-time Trump backer on Fox News, while Bianco is a former Oath Keeper).
Ethnic and geographical rivalries could matter too. Becerra, Calderon, and Villaraigosa are Latino; Yee is Asian-American; Thurmond is Black. Calderon, Porter, and Villaraigosa are from the greater Los Angeles area; Steyer, Swallwell, Thurmond, and Yee are from the San Francisco Bay area; and Becerra is from Sacramento. Schwarzenegger was the last California governor from Southern California, but he also represented the last gasp of truly moderate Republicanism.
While the field could shrink or expand even more before the filing deadline, the next governor of California probably won’t enter office with anything like the street cred and national prominence of the other 21st century chief executives, who often acted as though the state is an independent principality with its own foreign and domestic policies. Newsom will also leave some chronic fiscal problems, a perpetually fractious legislature, all sorts of natural resources and environmental challenges, and a housing “affordability” crisis that has spurred a national debate over a so-called “abundance” agenda prioritizing regulatory streamlining to speed up housing and other construction. It’s a lot, but whoever wins will become a lot more famous, fast.


The analysis on where Bushs’ gains came from was not bad, but it did not include a statement of methods. In other words, the reader is left to wonder by what the author means by “half is gains,” and so forth. Are these headings meant to express percentages of electorial or popular votes?
This is not suprising. During election night, one analyist after another talked about the importance of Bush running up the totals in Red States in order to make his election win seem more valid. Now we know how.
PS…the power of war to trump class predispositions is not exactly new..recall the Great Socialist Sell Out in France and Germany on the eve of WWI among many other examples
War unleashes all sorts of powerful counterintuitive emotions and behaviors and Bush made the very most out of it he possibly could have
In 2002, the Beltway Democrats defaulted the War on Terror and the War on Iraq to Bush hoping that in doing so, OUR issues, domestic issues, would come to into play.
Though the party by fits and starts came to appreciate the wisdom of Rove’s “hit em where they’re strongest”, it never fully shed the former mindset and two years later and indeed in the last two weeks of the campaign, felt the effects again.
Bush’s greatest achievement was turning the War on Terror away from a fact based debate on its conduct and turning it into a cultural values issue.
The IraQ war opened a window for reframing the War on Terror into a general camaign theme centered on Bush’s lies and incompetence.
Just as in the Bushevik slogan..The road to Jerusalem runs through Baghdad, so did the Demo road back to economic issues.
Lies and incompetence was a theme that the Democrats failed to pick up until very late in the game. IraQ and national security attacks brought the campaign out of its August to mid Sept slide but there was no hook at the ready to enable Kerry to benefit from the momentum gained….and he desperately neeeded one because the economy, if the econometric models suggest, just wasn’t weak enough to drag Bush down…
The determinants of this election were in place last year and dithering with our old mindset, we failed to act in a timely fashion to change the game
I live in North Jersey and there is no doubt that the Kerry margin of 7 poi nts compared to the Gore margin of 16 points was due to the fear of terrorism,25% of the 9-11 victims lived in NJ. Also the McGreevey scandal cut into the Kerry margin.When I look at the next 4 years,maybe its a blessing in disguise. The economy is going to come crashing down on Bush’s head in the next 4 years and a Pres.Kerry,unable to raise more revenue by a GOP Congress would have had the same problems Bush will have in the 2nd term.Would Kerry have turned out to be another Carter,a victim of economic circumstances beyond his control?Of course we will never know but something tells me the answer would have been Yes.
It doesn’t make me feel very optimistic for 08 to know that Bush had gains that were as across-the-map as this entry suggests. And widely distributed gains tend to dramatically undermine the voting fraud argument since not all states were doing e-voting.
For those pursuing the evote fraud angle, here is another website. The author is a computer scientist who has been warning of the risks since 2002. They are not strictly fraud risks but there are inherent limitations in the confidence one can have in any computer program.
http://www.notablesoftware.com/evote.html
1/2 + 1/3 + 1/5 is greater than 1. Yes, I know yours is a rough approximation.
How come you have not discussed the possibility of voting machine Fraud in trying to explain the discrepancy between exit polls and actual results.
Even if there was no fraud, should’nt we have audit trails for electronic voting machines? It is really not hard at all to have these machines print out paper receipts. How can we call ourselves a democracy if we cannot verify that votes are counted correctly?
Diebold will not provide the source code because it is a trade secret??? Does this pass the laugh test??? A sophomore in computer science can write a program to total the votes correctly!
I await the county-level analysis with great interest. I’ve been thinking about Ohio, and used the 2000 Census data to get my thoughts in order.
Columbus is the biggest city in Ohio, the 15th biggest in the country (in 2000), and has suburban-exurban counties that are among the fastest-growing in the country. Anecdotal evidence suggests that Columbus, unlike northern Ohio, is rather prosperous. Given all these factors, a detailed review of voting patterns in such an area would be very informative. Note that David Brooks claims that just such places are where the Dems “just don’t get it” and the Republicans do. (NYT, Nov. 6). Nothing like real evidence to check such a claim!
I am very puzzled by the apparent contradiction between the exit polls and county-level results. Exit polls show Bush doing better in cities and worse in rural and small-town areas than in 2000. The county-by-county maps published last week in the New York Times show Bush and Kerry each improving their percentages where they had already been strong, except in three areas: (1) Bush did better in the coastal strip from Delaware to Connecticut. (2) Kerry did better in the rural swath from Minnesota to northern Idaho, where religious affiliations are more Lutheran than evangelical and there is a long history of isolationism. (3) Kerry did better along the Quebec border where Bush’s French-bashing could hardly have pleased French-Canadian voters.
One question: Could Bush’s urban gains and Kerry’s rural gains be artifacts caused by reclassifications of areas between the 2000 and 2004 exit polls? As a result of the 2004 census, have exurban Republican areas been reclassified from rural to suburban, and sunbelt cities reclassified from less than 500,000 to more than 500,000? It seems unlikely that such reclassifications could cause apparent shifts of the magnitude shown in the exit poll, but the question needs to be looked at.
Ruy,
I live in Florida, and was surprised how “red” we became given the 2000 vote. I know there was a big GOTV effort here and the negative adds were omnipresent. Do you have any insight into any other factors that resulted in us being so red this time? BTW, we had a hotly contested Senatye race and the Repub, Mel Martinez, won by only a slim margin (<200,000 votes I think).
How do these figures compare to the population distribution? From your description, it sounds to me like Bush’s margin increased by about the same amount in all three regions (red, blue, purple), but especially so in certain states (NY, FL, TX, etc.).
For example, you say that a fifth of Bush’s gain came from the purple states. But don’t they have about a fifth of the population? Or am I missing your point?
As to point 1, Bush did not in fact win TN by 6 or more points in 2000. He did make huge gains in our state, but we became a solid red state this time, didn’t start out that way.
More Analysis please
Thanks for the analysis. I’d like you to look into the area where Kerry got more votes, I assume that there are some counties, and thus demographics where Bush lost?
Also I’d like to have a statistical analysis of the importance of Demographic factors in the vote, e.g. that the best fit to the Bush/Kerry vote is based on a correlation of the form of:
. 0.80*(numer of times per year goes to church)
. +
. 0.30*(Lives in a red county)
. –
. 0.25*(Earnings in $100.000)
Note: The numbers and factors are not true, but just ilustrative of the way I’d like to see the influance of the various factors that seem to effect/predict Bush/Kerry, and of course even better, I’d like to see the same for Bush/Gore and Dole/Clinton, so that we can see what the important factors in peoples vote really is.
http://pages.ivillage.com/americans4america/id17.html
Kerry’s considering unconceding and having a recount. He asked people to send firsthand experiences of disenfranchisement to his brother’s law office and his office is eagerly counting calls that are encouraging him to unconcede and ask for a recount! There has been massive evidence of voting fraud (see bottom for links) and there are 2 organizations you can support to uncover hard evidence of this fraud.
In This Post:
(1) Call/fax Kerry’s senate office, the DNC, and the Ohio Democratic Party.
(2) If you have first handexperience of voter disenfranchisement (not just articles) contact his brother.
(3) Support Two organizations that are uncovering hard evidence of fraud: blackboxvoting.org and votewatch.us
(4) Kerry can still request a recount in Ohio (and he may have the best chance of winning with a recount there)and perhaps elsewhere! He has until they count the provisional ballots 11-15 days after counting the provisionals. (It doesn’t matter if he’ll have a hostile Congress to work with because even if he can’t get much done as president at least he would prevent the havoc and destruction of 4 more years of W in which our rights, environment,economy, social security, and our very lives are at stake!) .
When I called they put me through to someone who asked for my state: they seem to be adding them up! Contact Kerry at (202) 224-2742 – Phone (202) 224-8525 – Fax
email form:
http://kerry.senate.gov/bandwidth/contact/email.html
and urge him to unconcede and do a recount in Ohio (and perhaps elsewhere) Also contact the DNC about this since pressure from them either way would influence Kerry. This is their phone number: 202-863-8000 This is the page for their email address http://www.democrats.org/contact/
(5) If you have witnessed or experienced disenfranchisement you can contact his brother’s law office–they are collecting this information which will be vital in considering unconceding at CKerry@Mintz.com (Don’t just email articles or they will be inundated with emails. They already know about the articles.)
(6) Help These Two Organizations Prove Fraud
There is anecdotal evidence of widespread fraud with the paperless voting machines. There are two groups working to uncover hard evidence who need your support.
(a) Please support the work of http://blackboxvoting.org– which is the only group uncovering hard evidence of fraud of the paperless electronic voting machines–with donations and/or volunteer work–they need to raise $50,000 to file freedom of info act requests for as quickly as possible to pay for records and the fees some states charge for them. If you can’t donate funds: http://www.eservicescorp.com/form.aspx?fID=912 ,
please donate time.
E-mail to join the Cleanup Crew. (they need all types from doing grunt work, to lawyers and programers) crew@blackboxvoting.org
(Please also contribute tovotewatch http://www.votewatch.us .
They need $250,000 to do a professional statistical analysis of the election which can be used as hard evidence.
This is an excellent organization that has been conservative in its approach, using highly respected statisticians and developing trusted relationships with key media contacts. They are collecting and analyzing data to determine if there was fraud in the election as seems to be indicated by the 5% (or so) discrepancy between exit polls and reported results from the touch screen voting with no paper trail vs. the other types of voting where exit polls closely matched reported results.
If you decide to move forward with a tax-deductible contribution, please make your check payable to Votewatch (ID# 94-3255070) and send it to:
Votewatch
c/o: The San Francisco Foundation Community Initiative Foundation
(SFFCIF)
Attn: David Barlow
225 Bush Street
Suite 500
San Francisco, CA 94104
This is an excellent organization that has been conservative in its approach, using highly respected statisticians and developing trusted relationships with key media contacts. They are collecting and analyzing data to determine if there was fraud in the election as seems to be indicated by the 5% (or so) discrepancy between exit polls and reported results from the touch screen voting with no paper trail vs. the other types of voting where exit polls closely matched reported results.
If you decide to move forward with a tax-deductible contribution, please make your check payable to Votewatch (ID# 94-3255070) and send it to:
Votewatch
c/o: The San Francisco Foundation Community Initiative Foundation
(SFFCIF)
Attn: David Barlow
225 Bush Street
Some sites with voterfraud info:
http://www.stolenelection2004.com
http://pages.ivillage.com/americans4america/id17.html
[URL=http://radtimes.blogspot.com http://www.whatreallyhappened.com/2004votefraud.html%5Dhttp://radtimes.blogspot.com http://www.whatreallyhappened.com/2004votefraud.html%5B/URL%5D
http://legitgov.org
http://democrats.com
250% of vote margin came from Karl Rove’s Bat Cave