Articles and Analysis


New Hampshire: So What Happened?

Topics: 2008 , New Hampshire , The 2008 Race

There is obviously one and only one topic on the minds of those who follow polls today. What happened in New Hampshire? Why did every poll fail to predict Hillary Clinton's victory?

Let's begin by acknowledging the obvious. There is a problem here. Even if the discrepancy between the last polls and the results turns out to be about a big last minute shift to Hillary Clinton that the polls somehow missed (and that certainly sounds like a strong possibility), just about every consumer of the polling data got the impression that a Barack Obama victory was inevitable. One way or another, that's a problem.

For the best summary of the error itself, I highly recommend the graphics and summary Charles Franklin posted earlier today. Here's a highlight of how the result compared to our trend estimates:

What we see for the Democrats is quite stunning. The polls actually spread very evenly around the actual Obama vote. Whatever went wrong, it was NOT an overestimate of Obama's support. The standard trend estimate for Obama was 36.7%, the sensitive estimate was 39.0% and the last five poll average was 38.4%, all reasonably close to his actual 36.4%.

It is the Clinton vote that was massively underestimated . . .Clinton's trend estimate was 30.4%, with the sensitive estimate even worse at 29.9% and the 5 poll average at 31.0% compared to her actual vote of 39.1%.

So what went wrong? We certainly have no shortage of theories. See Ambinder, Halperin, Kaus, and, for the conspiratorially minded, Friedman. The pollsters that have weighed in so far (that I've seen at least) are ABC's Gary Langer (also on video), Gallup's Frank Newport, Scott Rasmussen and John Zogby. Also, Nancy Mathiowetz, president of the American Association for Public Opinion Research (AAPOR) has blogged her thoughts on Huffington Post.

Figuring out what happened and sorting through the possibilities is obviously a much bigger task than one blog post the morning after the election. But let me quickly review some of the more plausible or widely repeated theories and review what hard evidence we have, for the moment, regarding each.

1) A last minute shift? - Perhaps the polls had things about "right" as of the rolling snapshot taken from Saturday to Monday, but missed a final swing to Hillary Clinton that occurred over the last 24 hours and even as voters made their final decisions in the voting booth. After all, we knew that a big chunk of the Democratic electorate remained uncertain and conflicted, with strong positive impressions of all three Democratic front-runners. The final CNN/WMUR/UNH poll showed 21% of the Democrats "still trying to decide" which candidate they would support, and the exit poll showed 17% reported deciding on Election Day with another 21% deciding within the last three days. Polls showed Clinton polling in the mid to upper 30s during the late fall and early winter before a decline in December. Perhaps some supporters simply came home in the final hours of the campaign.

I did a quick comparison late last night of the crosstabs from the exit polls and final CNN/WMUR/UNH survey. Clinton's gains looked greatest among women and college educated voters. That pattern, if it also holds for other polls (a big if) seems suggestive of a late shift tied to the intense focus on Clinton's passionate and emotional remarks, especially over the last 24 hours of the campaign.

2) Too Many Independents? - One popular theory is that polls over-sampled independent voters who ultimately opted for a Republican ballot to vote for John McCain. I have not yet seen any hard turnout data on independents from the New Hampshire Secretary of State, but the exit poll data does not offer promising data for this theory. As I blogged yesterday, final Democratic polls put the percentage of registered independents (technically "undeclared" voters) at between 26% and 44% (on four polls that released the results of a party registration question). The exit poll reported the registered independent number as 42%, with another 6% reporting they were new registrants. So if anything polls may have had the independent share among Democrats too high.

On Republican samples, pre-election pollsters reported the registered independent numbers ranging between 21% and 34%. The exit poll put it at 34%, with 5% previously unregistered. So here too, the percentage of independents may have been too low.

Apply those percentages to the actual turnout, do a little math, and you get an estimate of how the undeclared voters split: roughly 60% took a Democratic ballot and 40% a Republican. That is precisely the split that CNN/WMUR/UNH found in their last poll although other

Keep in mind that the overall turnout was over 526,671 (or 53.3% of eligible adults). Eight years ago (the last time both parties had contested primaries) it was 396,385 (or 44.4% of eligible adults at the time). That helps explain why we may have seen an increase in independents in both parties.

Of course, we are missing a lot of data here: Nothing yet on undeclared voter participation from the Secretary of State, and roughly half the pollsters never released a result for party registration.

3) Wrong Likely Voters? OK, so maybe they had the independent share right, but perhaps pollsters still sampled the wrong "likely voters" by some other measure. The turnout above means that pollsters had to try to select (or model) a likely electorate that amounted to roughly half the adults in New Hampshire, they reached with a random digit dial sample.

Getting the right mix is always challenging, possibly more so because the Democratic turnout was so much higher than in previous elections. That's an argument blogged today by Allan McCutcheon of Edison Research:

In 2004, a (then) record of 219,787 voters turned out to vote--the previous record for the Democratic primary was in 1992, when 167, 819 voters participated. This year, a record shattering 287,849 voters participated in the New Hampshire Democratic primary--including nearly two thirds (66.3%) of the state's registered Democrats (up from 43.3% in 2004). Simply stated, the 2008 New Hampshire Democratic primary had a voter turnout rate that resembled a November presidential election, not a usual party primary, and the likely voter models for the polling organizations were focused on a primary--this time, that simply did not work.

One way to assess whether polls sampled the wrong kinds of voters would be to look carefully at their demographics (gender, age, education, region) and see how they compared to the exit poll and vote return data. Unfortunately, as is so often the case, only a handful of New Hampshire pollsters reported demographic composition.

4) The Bradley/Wilder effect? The term, as wikipedia tells us, derives from the 1982 gubernatorial campaign of Tom Bradley, then the long time African-American mayor of Los Angeles. Bradley led in pre-election polls but lost narrowly. A similar effect, in which polls understated the support for the opponents of African-American candidates seemed to hold in various instances during the 1980s. Consider this summary of polls compiled by the Pew Research Center for a 1998 report: which they updated in February 2007:


Note that, in almost every instance, the polls were generally about right in the percentage estimate for African-American candidate but tended to underestimate the percentage won by their white opponents. The theory is that some respondents are reluctant to share an opinion that might create "social discomfort" between the respondent and the interviewer, such as telling a stranger on the telephone that you intend to oppose an African-American candidate.

Of course, the Pew Center also looked at six races for Senate and Governor in 2006 that featured an African-American candidate and did not see a similar effect. Also keep in mind that that all of the reports mentioned above that show the effect were from general election contests, not primaries.

What other evidence might suggest the Bradley/Wilder effect operating in New Hampshire in 2008? We might want to consider whether the race of interviewer or the use of an automated (interviewer-free) methodology would have an effect, although these kinds of analyses are difficult, because other variables can confound the analysis. For what it's worth, the final Rasmussen automated survey had Obama leading by seven points (37% to 30%), roughly the same margin as the other pollsters. We might also look at whether pushing undecided voters harder helped Clinton more than other candidates.

Update: My colleagues at AAPOR have made three relevant articles from Public Opinion Quarterly available to non-subscribers on the AAPOR web site.

5) Non-response bias? We would be crazy to rule it out, since even the best surveys are getting response rates in the low twenty percent range. If Clinton supporters were less willing to be interviewed last weekend than Obama supporters, it might contribute to the error. Unfortunately, it is next to impossible to investigate, since we have little or no data on the non-respondents. However, if pollsters were willing to be completely transparent, we might compare the results among those with relatively high response rates to those with lower rates. We might also check to see if response rates declined significantly over the final weekend.

6) Ballot Placement? Gary Langer's review points to a theory offered by University Prof. Jon Krosnick, that Clinton's placement near the top of the New Hampshire ballot boosted her vote total. Krosnick believes that ballot order netted Clinton "at least 3 percent more votes than Obama."

7) Weekend Interviewing? I blogged my concerns on Sunday. Hard data on whether this might be a factor are difficult to come by, but it is certainly an issue worth pursuing.

8) Fraud? As Marc Ambinder puts it, some are ready to believe "[t]here was a conspiracy, somehow, because pre-election polls are just so much more valid than actual vote counts." Put me down as dubious, but Brad Friedman's Brad Blog has the relevant Diebold connections for those who are interested.

Again, no one should interpret any of the above as the last word on what happened in New Hampshire. Most of these theories deserve more scrutiny and I agree with Gary Langer that "it is incumbent on us - and particularly on the producers of the New Hampshire pre-election polls - to look at the data, and to look closely, and to do it without prejudging." This is just a quick review, offering what information is most easily accessible. I am certain I will have more to say about this in coming days.



Mark, when do we hear from Margie Omero on the women's vote? Tell her to hurry up. :)


Sean Truly:

Wtf,Hillary won asshole,So stop wining about it



Mark, I asked several related questions in the comments below Charles Franklin's excellent posting on the pre-primary polls that I hope you and he will consider addressing. Now I've got two more. One, did the exit polling consortium reweight the data after the polls closed to reflect the actual results? And two, can you tell us whether the research on voting and race finds any gender or education differences for things like the Bradley Effect?


richard Pollara:

The biggest shift in voting between Iowa and NH was among women. In Iowa where you had to publically declare your vote Obama won the female vote by several percentage points. In NH with secret voting, Clinton won the women's vote by 13% a shift of almost 17%. Isn't it possible that women, in the privacy of the voting booth, voted their hearts? They didn't tell their husbands and they didn't tell the exit pollers. I think it is a far more likely scenario than the Bradley Theory.


Michael Xia:

I'd like to argue against the notion that there wasn't an overestimation of Obama's support. It's true that the raw number matches with what the polls showed, but remember all the polls have undecided as an option, so they were not normalized. If we normalize the poll data against those who were decided, the projected Obama support would be in fact above 40%. So either all the undecided voters went with Hillary, a notion not supported by exit polls, or there was a clear shift from Obama's camp to Hillary's during the last 24 hours, or there was some intrinsic bias the way the polls were conducted (whether it's Hilary effect or Bradley effect or something else).


paul lukasiak:

Well, I looked at the last suffolk poll internals (Jan 6-7, 500 LV), and what did i find?


That if you examine the internals, you got a whole different picture of the race than if you just looked at the "Headline Obama 39, Cliton 34" data.

What you found, when you looked at the answers to the question "How likely are you to change your vote? Very Likely, Somewhat Likely, Undecided, Unlikely", (Banner 2, Q3) was that only 74% of people who planned to vote in the Dem primary were "unlikely" to change their minds. And that Hillary and Obama were vitually tied among voters unlikely to change their minds (N=371, Hillary 146, Obama 153).

Obama had almost twice the number of "at risk of changing" (N=129) than hillary (Obama 41 to hillary's 23).

Moreover, and contrary to popular wisdom, women stuck with their choices, and the last day "surge" to Clinton was from men who changed their minds, and and a higher than predicted relative turnout from women.

The suffolk poll shows that men were more likely than women to change their minds. Men made of 56% of the "changeable" sample, but 47% of the total sample, women made up 53% of the total sample, but only 44% of the "changeable" sample.

And with the data showing Clinton doing better than Obama among women (46% Clinton, 33% Obama), and significantly worse among men than Obama (20% Clinton, 46% Obama) the fact that a higher percentage of women than men were unlikely to change their vote worked to Clinton's advantage.

In fact, women do not appear to have changed their minds at all. Clinton got the same percentage of women's votes in both the Suffolk poll and the exit polls (46%), while Obama increased his percentage of the female vote by 1% (to 34%)

Things were entirely different on the male side of the ledger, where Clinton went from 20% in Suffolk to 29% in the exit poll when it came to men, while Obama lost 4% (to 42%).

And because Clinton held her own relative to Obama for the women's vote, the higher than predicted percentage of women voters worked to Clinton's advantage. (Suffolk 53% women, exit poll 57% women.)

In sum, the suffolk pole predicted a highly volatile dead even race that could have gone either way, but the nature of the volativity favored Clinton.



I agree. In order to accept the Bradley/Wilder effect, you would have to suppose 2 things:

1) Many people allowed biggotry to effect their vote, but NONE of them chose to vote for the only viable White Male in the race (John Edwards), who balloted right at or below his poll numbers.

2) Only women are subject to the Bradley effect, while men (who voted for Obama by an 11 point margin- roughly where it was expected to be) are totally immune.

I find BOTH of these thing implausable, and you would have to believe BOTH for there to be a Bradley effect. More likely, I think, the Clinton ground machine in NH was much better than people thought, combined with Hillary dominating 2 days of news cycles by first crying, and then having Bill rip into Obama. Toss in NH's penchant for not wanting to end the race prematurely, and you have your answer.

So who will win? I have a feeling African American Females will get to decide, starting in South Carolina. If this renews their doubts about Obama's viability, then She wins, but if their not-so-subtle race-baiting (Spade Work? Really?) makes them stauch Obama Girls, or if they merely continue to believe he is viable in a general election, then He does.



Another possibility might be that Clinton just had that much stronger of a ground game in NH than Obama did - including absentee ballots that may have been cast weeks ago. I have no idea how each did at getting their supporters to vote absentee, but given Clinton's longstanding lead in the state, it wouldn't be surprising to find that she built up several thousand votes (maybe even a few thousand over Obama?) before Iowa even happened. That still wouldn't be enough to make up for the 10-15 point difference between the polls and the results, but it could be a big part of the story.



The more analysis I hear and read, the more I agree with what Richard suggests... the Bradley effect feels counterintuitive. My gut says its a "Hillary effect"- the rampant sexism in the media and on the internet and the chest-thumping nastiness of many Obama supporters made it seem socially unacceptable to admit to voting for Hillary / a female candidate. Women are socialized to not be publicly disagreeable and to preserve male "ego." What they do in the privacy of the voting booth is another story. I'd love to know if there is anything to back this up or if I'm just sleep-deprived.

Sean Truly: "stop wining"? you must mean "winning" - and that, dear, is the whole point.


Arghh. I was praying that I wouldn't have to do my own research to check out the absentee ballot theory (i.e., that HRC benefitted from robust pre-Iowa voting). But no such luck. Is this really so implausible that it doesn't make your long list of good and bad theories? Am I hallucinating, or didn't we all decide that early voting had become a major problem for pollsters some time ago? What am I missing? Doesn't the theory explain a big chunk of the disparities?

Ed Kilgore



paul lukasiak - that was fascinating...



ed, of course you have to check the absentee ballot numbers... or if you find a link to them, I'll do it! ;)

Ultimately, I don't think you are going to find anything significant, however. I think that the "error" was caused by an incorrect estimation in the relative percentage of men and women who voter. "Gender" is usually one of those variables that pollers use to make sure that the sample is correct -- and with such a huge differences between how the genders voted, (women +14 clinton, men +13 obama), if you decided to use the 2004 primary distribution of 54% women, 46% men, you were gonna be off by a considerable margin to begin with.



Thank you for the Suffolk crosstabs. We now have three, CNN, ARG, and Suffolk. CNN and ARG show the strong late shift being mostly among women, while Suffolk showed it mostly among men. We clearly need to see all the crosstabs just so we can tell whether Suffolk (which got the GOP race very wrong) was an outlier, or whether this is something which bounces back and forth between all the polls.




I have the same question as cmmcd. Was the exit polling re-weighted to account for the results. I can't seem to get a clear answer on this anywhere. Please help.



To rule out fraud, far greater transparency into the vote-processing mechanisms is necessary. Currently, there is ample cause for suspicion, and a full investigation and recount is warranted given the aberrant results. For such a confluence of highly fortuitous events to occur simultaneously and independently to propel Clinton over the top seems implausible, to say the least. Although it is to early to draw such drastic conclusions, other explanations don't adequately explain the issue given the magnitude of the pro-Hillary swing.


Mark Blumenthal:

ccmd and Mike: The network exit poll results that we see online are always weighted to bring the overall result into line with the actual count. That's standard practice, and something I explained in detail here. The one and only place that the results are not similarly reweighted is for the Iowa Caucuses.


Mike Rappeport:

The problem with the fraud argument is that the exit polls (which are not only the most accurate polls since they have much less of a sampling problem, but are also done by a completely independent set of people) corroborate the actual results. Thus unless there was somehow the same fraud in the exit polls as in the actual vote counts, fraud is simply contrary to the few "facts" available.


Joe From the Beach:

You got it wrong Rappeport. The Exit polls show an Obama win http://www.latimes.com/la-exitpoll-nh-graphic,0,7161708.htmlstory?coll=la-home-center

as did the nine pre election polls.

What no one wants to say is that Dieblod hackable machines where used in coming up with the Clinton win. http://www.globalresearch.ca/index.php?context=va&aid=7753

Fraud is the only logical answer

I fear for the Republic



Regarding the fraud hypothesis:

On BradBlog, they link to this page -- http://ronrox.com/paulstats.php?party=DEMOCRATS -- which has the vote totals by township, along with the voting method (electronic or paper).

One basic question is whether paper-voting towns produced different results from electronic-voting towns. Of course, town size correlates strongly both with voting method and the Clinton/Obama ratio. So I took the numbers above and did basic matching using the only data I had, total votes, which I presumed correlated with town size, and thus (hopefully) with other important demographic characteristics. That is, I took the 91 towns that voted electronically and matched each one (using matchit in R) with a similarly sized paper-voting town, and then compared the vote percentages for Clinton and Obama in those two populations. The results?

....................Cl ...... Ed ...... Ob
Electronic.... 39.2%..17.5%..35.8%
Paper ......... 38.5%..18.0%..36.1

Ie, the two voting types seem to have produced nearly identical results. Of course, more demographic data to match on would be nice, but I think this puts a big burden of proof on the doubters.


Nonresponse bias is likely, I think; and there *are* ways to investigate it. Who had more 'soft' support? Clinton. Whose numbers were seriously worng? Clinton's. Who are less likely to stay on the phone? Soft supporters

One could also sample a bunch of supporters of different candidates and measure degree of support.



I feel there was just an inncorrect data for the polls concerning the problem with the fraud argument is that the exit polls, which are not only the most accurate polls since they have much less of a sampling problem, but are also done by a completely independent set of people. Also,I would rather have a woman in the white house with proven political background, than ever have a Republican in the White House That Supports the American public being raped by oil companys......They have made Billions since President Bush has been in the White House. Also, Obama, would not be a bad choice if he had more political experience. I support Hillary Clinton, she can do a fine job as President. She has the political experience needed to do the job.......GO HILLARY!!!!


(Ed asked this question several places, so I'm answering again here.)

Ed Kilgore asks above about absentee voters in New Hampshire. Here is a first cut answer. (And no, Ed, I don't think you are a crank at all.)

New Hampshire requires you to certify you will be out of town or are physically unable to vote in person, or have a religious reason for not voting in person on election day. It is a misdemeanor to vote absentee unless these conditions apply.

So there is no early voting and absentee ballots are the traditionally restrictive type, not the more recent "easy" absentee ballots.

The result, in 2004, was that 4.99% of ballots statewide were cast by absentee vote for the Democratic primary.

That is not enough to account for the differences we see between polls and outcome, even granting (and I do) Ed's point that absentee votes would have been cast before Iowa. With only 5% absentee, the absentee vote would have to be 20 points different from the election day vote to move the total vote by just ONE Percentage point. For example, 30%*.95+50%*.05=31%.

The proportion of absentee votes have not yet been released by the New Hampshire Secretary of State's Office. Because the absentee rules have not changed, I don't expect a dramatic increase in absentee participation in 2008. Even if absentee votes doubled, the proportion seems still too small to account for the 8 point underestimate of the Clinton vote.

Thanks for raising this. In many states it might be much more important. Just not in this one.




Thanks, Mark, for directing our attention to the details of the NEP weighting. I will tackle the Pew report you cite to see if it addresses gender and education differences, but it appears that only a careful study of yesterday's NH exit data weighted for turnout but not for voting results would enable researchers to tease out a possible Bradley and/or Hillary effect (will "Hillary Effect" be a new pollster term of art after this?).

Thanks to everyone who has taken the time so far to look more closely at the internals of both types of polls and offer their analysis. It is obviously even more difficult and sensitive for pollsters to tackle questions about possible racist and sexist voting behavior than the usual disagreements about methodology.

But if older white women (in this case, Democrats and Independents) are more prejudiced against black candidates and/or more likely to vote for a woman than white men are, that would seem particularly important to know about in a dynamic environment where the news media's access to pre-election and exit poll data shapes the political expectations for candidates in a multi-contest race like the presidential nominating process.

By taking the pre-primary and exit polling at face value the media may have set up unrealistic expectations for the outcome of a contest that included both a white woman and a black man. At least one report I saw said that the Obama campaign regularly discounts his support in their own internal polling to account for the Bradley effect. Perhaps Mark Penn will comment about whether he does so for the Clinton campaign.



Can I offer another theory on this.

Hillary is always perceived as a polarizing figure in American politics. Could you be possible that some of her supporters, especially female simply are reluctant to tell pollster they will vote for her due to peer pressure in the household. Let's see this female voter's husband is not a big Hillary fan in the family, certainly his wife does not want to talk to pollster or lie to the pollster.

Since Iowa is a caucus, not primary, the polls might be more accurate under such scenario.

Some anecdotal evidence. Clinton's two senator race, especially 2000, she seems to be doing much better than final polls suggested. The polling vs. actual discrepancy in 2006 seemed to be minor, it might be because she has established her stature in New York, and some of her supporters no longer felt the peer pressure.

You can probably compare the history of Clinton's polling history vs. actual.


Matthew, Roanoke VA:

I really don't see the difficulty in grasping what happened.

A huge number of women, that will vote for Hillary because they want to see a woman president, were suddenly confronted by the stories that Hillary was going to pull out of the race if she didn't win NH.

Many won't admit that her gender is the reason for her support since that makes as much sense as voting for someone because he is a white male.

Before the rumor of her possibly pulling out, many, I believe, were planning on doing something else with their time than actually voting. They did not see the importance of this one primary when they thought Hillary would continue in spite of losing NH, and eventually secure the nomination. The story of her pulling out really woke them up.


Steve D:

@ Joe From the Beach

Regarding that LA Times exit poll you linked to above. Unless my math skills are horrible, (quite possible) it shows Hillary in the lead.

Like all the other exit polls I've looked at this evening, the LA Times info doesn't give a straight breakdown as to who was leading. Everything is broken down by gender or some other variable.

I re-calculated a few of the variables to find the straight unbroken down figures. Curiously, none of the variables (gender, college, party id) provided exactly the same total vote result. But generally it's Hillary with 38.x and Obama with 36.x to 37.x.

Of course, this like the other exit polls have been weighed in deference to the actual result.

All that said, a number of reporters have stated that the exit poll numbers they were shown last evening all put OBama in the lead.

Are the unadjusted exit poll numbers available anywhere?



. We clearly need to see all the crosstabs just so we can tell whether Suffolk (which got the GOP race very wrong)

wow, you're right.

It looks like Suffolk completely missed McCain's female support (20% in Suffolk, 38% in exit poll.)

The only explanation that comes close to accounting for Suffolk's performance here is that they have the GOP race being even more volatile than the Dems -- (only 67% were unlikely to change their vote)... amd more remarkably, only 59% of their female GOP sample said that they were unlikely to change.

And while its still possible to account for the huge discrepancy in women voting for McCain by saying that 18% of women changed their minds and voted for McCain, this is the first I've heard of that.


Yeah, joe_from_the_beach got it wrong, those are adjusted exit polls which show a Hillary win.

I am not sure that the machine-counted fraud can be ruled out by the evidence we have. That is the beauty of those paper ballots - recounting a few townships by hand would be quick and simple to see if there are any variances.


Joe From the Beach:

The La Times exit poll information was the best I could find. Please note the first question. 44% thought Obama was the most likely candidate to beat the Republican candidate in November. Clinton got only 35%. So people favored in Clinton in the actual vote even though they thought she would be the weaker candidate. Not probable.

The more straightforward question: Who did you vote for, does not appear to have been asked or tabulated.

There must be considerable pressure on the media to down play any allegation of voter fraud. As a practical matter, people may die unnecessarily, if the world perceives the US as having dishonest elections. The correct approach here, however, is not denial but that US elections must be squeaky clean and beyond even the appearance of possible of fraud.

The counter to this concept is the electronic machines. The particular Diebold machine can in hacked in seconds, as the software has a Windows base. Historically the Diebold machines extensive use came on with the incredibly indiscreet assertion of Wally Waldron then Diebold's CEO, that he would deliver Ohio for Bush in 2004. Next came the 18,000+ under vote in the Sarasota Congressional race. (Not Diebold but another similar flawed machine) Now we have Clinton's miracle win.

This must be cleaned up and the bets way is to abandon the machines, until if and when they cannot be hacked and every vote counted correctly


Randy Brooks:

I was there on the ground working for McCain and I am a Market Research professional with 35 years of experience.

The news in New Hampshire from Monday on was "will Hillary stay in the race after her loss in New Hampshire". It was a done deal, over.

I think Obama's people and his supporters probably relaxed and did not turn out their vote. The Hillary "tears" episode clearly helped her and the ballot that does not rotate positions is shocking. All of this adds up to what happend.



This on-the-ground report seems to partially support Hillary's 'hidden women' votes theory...


Erica was the first to notice something odd

Just about every other woman would give us the thumbs up as they passed. Understand, we heard the polls too and had convinced ourselves that we'd be losing this day. But they not only gave us thumbs up, most of them were doing it in a shy surreptitious manner as if they didn't want their husbands to see. This continued all day peppered with the winks and loud Go Hillary we got from many men.


Fred P:

My thoughts:

1) A last minute shift?
Are there any exit polls that bear this out? The ones I've seen that look at when the voter made up their mind can only make up for a tiny fraction of the roughly 10-point discrepancy (1-2%).

2) Too Many Independents?/3) Wrong Likely Voters?
These seem the most likely to me, largely because I'm not finding contradictory evidence.

8) Fraud?
Again, one would expect this to be spotted via exit poll; a 10-percent gap between votes and reported vote is pretty obvious. I haven't seen an exit poll that is showing a large deviation from actual results.


Mark Lindeman:

Fred, I think the exit poll "when decided" table (although it shouldn't be taken literally) does bear out a last-minute shift. You have to compare the numbers to what they should have been if the Obama bounce had held up. In 2004, people who said they had decided in the last week (two categories combined) gave Kerry a 19-point margin, compared with 6 points among the rest.

The 2008 when-decided table tends to indicate either that there never was an Obama bounce, or that it unbounced.



Why wouldn't fraud be the number 1 explanation? Its the one with the most historical support for explaining such an anomalous result. Ballot placement explains only less than 3% and there's as much as 16% variance in the polls vs results (and some commenters have said that it was illegal to have Clinton #2 on all ballots per NH Supreme Court decision).


Steve D :

I think Bradley effect is much more likely than fraud. MUCH more likely.

We've never seen this level of provable fraud in a recent US vote. But we have seen provable instances of the Bradley effect. When Bradley has happened in the past, it has caused exactly this sort of polling error.

Many in the media have reported that the exit numbers on election evening showed an Obama win. Last night on MSNBC, Matthews said a precinct they'd analyzed showed bradley effect voters.

From Matthews comments, I gather someone with access to the raw numbers is doing this analysis right now.

Does anyone know when this raw-number analysis will be released?



It is inconceivable to believe this election result is accurate.
1. The constant chant about a flurry of women in every last little town running to respond to Hillary's whining and blubbering is b.s. There may have been an increase in some within the Democratic party -but I'm a woman and resented her show of self pity. And that would hardly help non-Democratic Independent women rally to her.
2. And the few extra women would hardly reach so far over and beyond the Independent voters supporting Obama. We're talking a 15 point difference.
3. More importantly -look at the Republican side. It matched the polls. Look at the Democrats hand counted votes -in every last town using hand counted methods the votes matched the polls. Obama won 83 of those towns to Clintons 45 -just 7 points under a two to one ratio for Obama. In a big turnout he was expected to win big.
But now watch the count as it added up -not once did it ever show Obama in the lead or fluctuating as you would expect to see in a close race
Now look at this to find your answer:
LHS -already known for it's corruptible optical scan machines -counts all of New Hampshire machines and counts 81% of the vote.
Their results??
59% to Clinton - 38% to Obama
Almost an exact reversal of the wide lead Obama held in the hand counted towns.
And also the #1 trick used in memory card corruption. Obama vote totals are reversed with Clinton -and that's why every hour of the tally consistently showed Hillary in the lead.

My point is not to change the outcome -but to push you pollsters to make a serious effort here to catch the real criminals once and for all and to stop this bullsh**.
If you value your own jobs, you'll find and declare to the world the real truth of what happened in the New Hampshire election of January 2008.


Dan G:

#1 seems like the most plausible explanation -- but it shows the polls weren't necessarily wrong.

The polls reported that, shortly before the election, Obama led 37-30, and that 20% of voters remained uncommittted. It might well be that that was an entirely accurate snapshot.

The press viewed that snapshot and leapt to the conclusion that Obama would win. That leap was premised on the assumption that unaffiliated voters would break evenly. And that was a faulty assumption - they broke more heavily for Hillary.

The polls weren't "wrong" any more than I would be wrong to tell you the Giants lead 28-23 with ten minutes left in the game. When you turn around and tell everyone the Giants will beat the Patriots, you're making some very significant assumptions about future trends, and you're running the risk of being wrong.

If the polls showed there were few undecideds, that would be one thing. But they didn't. The fault was not with the pollsters but with the pundits who should have noticed the high number of undecideds and cautioned that the race was still up for grabs.



The La Times exit poll information was the best I could find. Please note the first question. 44% thought Obama was the most likely candidate to beat the Republican candidate in November. Clinton got only 35%. So people favored in Clinton in the actual vote even though they thought she would be the weaker candidate. Not probable.

in fact, its highly likely.

Look at the Suffolk poll (q6 banner2) which asked the "most likely to be the next president" question

(note that potential responses included Republicans, and "undecided")

It came out with similar numbers...25$ said Hillary, and 38% said Obama.

61% of Obama voters said that he would be the next president, while 11% said Hillary. 57% of Hillary voters said Hillary, while 19% said Obama. And only 7% of Edwards voters said Hillary, while 27% said Obama (and only 29% said Edwards).

6% of the Democratic sample thought a Republican would be president (including 6% of Obama supporters who though McCain specifically would be president.) And 20% of the GOP sample thought one of the Democrats would be the President.

Now, while the questions aren't exactly the same, they show that people's choices about who they vote for, and their opinions of the political landscape, are separate and distinct.

And if you look at the exact exit same poll that you cite to claim that the numbers don't make sense, you'll note that only 6% of Democrats thought Electability is the most important quality in a candidate.

Personally, I think regardless who the democratic nominee is, that person will wind up being president. But I think Edwards has the best chance of winning against a Republican, I still think Hillary will be the nominee.

The more straightforward question: Who did you vote for, does not appear to have been asked or tabulated.

um, yes it was. its just that they didn't bother to make that a separate question on the web page -- its all about breaking down each candidates support.



The press viewed that snapshot and leapt to the conclusion that Obama would win. That leap was premised on the assumption that unaffiliated voters would break evenly. And that was a faulty assumption - they broke more heavily for Hillary.

actually, there were precious few "unaffiliated" voters at the end. What there was were a lot of "leaners".

One of the problems here is that there were lots of people who had still not made up their minds for certain, but the pollster's wanted an answer anyway. Suffolk asked voters "Of these, the major eight candidates listed alphabetically on your ballot, are...for whom will you vote or toward whom would you LEAN at this time?", then followed up to find out how likely (very, somewhat or un-) to change their minds.

Other pollsters would ask "who are you voting for" and if the answer was "I don't know for sure for", they would be ask "well, are you leaning toward anyone", and the answer to that question was almost invariably "yes", and the answer was added to the candidate's totals.

One of the big problems is that polls report "leaners" as if they are committed to a candidate in their "Headline" data...and play down the existence of "leaners" if they mention them at all. And because most pollsters don't let their crosstabs out to the public (at least, not without paying for it) its practically impossible to understand what each poll really represents.

One of the things I keep pointing out is that in the suffolk poll, among "committed" voters, Obama had only a 1.5% lead --- and a significantly higher percentage of Hillary's voters were committed than Obama's.

Given the lateness of the Obama surge (and the Edwards surglet), these "uncommitted" voters were likely to be people who had planned on voting for Hillary, and were having second thoughts -- some were still leaning toward Hillary, while others had switched to leaning toward Obama (or Edwards -- only 68% of Edwards voters were unlikely to change their minds). And I'd further like to posit that Obama didn't close the sale in the last days of the campaign, and Hillary's "changeable" voters stayed with her, while a big chuck of Obama's "changeable" voters went back to Hillary.


Gary Kilbride:

This is obviously a numbers oriented website and I'm typically numbers oriented. But in 20+ years in Las Vegas I learned early to buffer the numbers reliance with simple common sense.

In this case the common sense always screamed a huge bump for Obama out of Iowa was not going to happen. That's why I'm still amazed at the scrambling to rationalize the result.

The theme for more than a year is Hillary is near 100% name recognition level, with heated opinions both ways, more or less locked in place. So how in Monty Python lunacy does that equate to a massive shift either way, based on one result? It's laughable contradiction. You can't blindly embrace historical trends without applying current situational realities.

There's no way you can deny she has always been the frontrunner, with considerable support among women, the majority voting block. Also no way to deny a huge chunk doesn't want her. So by definition I always felt she was immunized from a typical bounce scenario, in either direction.

Due to a family matter I wasn't paying attention to polling for several weeks prior to NH, and tuned in that night expecting a tight race with a likely Hillary win. That's what screamed sensible, based on Obama winning Iowa, earning a small bounce, and with the background knowledge of Hillary's long term significant advantage in NH. I'll always be astonished this is viewed as some sort of landmark upset.


Fred P:

@Mark Lindeman-

Good point; I'll go over the data again with that in mind.



Still What happened!

As I watched Morning Joe this morning still trying to find out what what wrong with the polls it dawned on me. Barack had speeches with 1500 people at then Monday perhaps they didn't like his message! Maybe they thought he couldn't deliver on his comparison to JFK he likes to make. Perhaps they "after seeing him speak " decided he was a bag of wind and in no way could he deliver on this dream he is living. The voters of New Hampshire decided he like most all politicians, was trying to sell them a bunch of "lies"
"fairy tails""dreams" or Band Instruments for a Grand Parade" just to get elected. How about that as an explanation.

Quick Ramblings



I have a suspicion. And it's just that (I'd love to know ow to confirm this).

What if most NH Democrats are quite confident in the candidates they have in front of them right now? From just a cursory skim of most blogs, polls and other sources, most Democrats don't seem especially upset about their choices (which may be unusual).

Going from that, I can easily see how a certain calculus may take place. It seems plausible to me that since the voters have a lot of confidence in the potential candidates (Democrats, not Republicans) that changing their minds quickly based on how the press deals with a candidate would be easy.

If I don't hate Hillary or Obama (and I think most Democrats like both of them) and Hillary is the underdog going into the primary, why not just vote for her?



Sam Dixon:

Hopefully you don't mind input from "damn foreigners up in Canada" but the results are an expectable vulnerability given the support bases of these two candidates.
Obama's support has come from a combination of historically marginalized vote (Hispanic), trending declining vote (youth), and low commitment vote (independents).In other words people still in the process of engagement. Clinton's support is from the party faithful and the establishment.
As entertaining as conspiracy theories may be the results are indicative of the difference between involvement and commitment; Obama's support is soft support that will be an on-going challenge to maintain because these are "people outside the system" while Clinton's support are old hands at this. The "Get Out The Vote" strategies are very different so New Hampshire isn't an upset it's a learning oppourtunity - if Obama's people do then Clinton has a fight on her hands;if Obama's people don't then expect rapid decline.


Joe from the Beach:

Hold on before you reach any conclusions on this election. CounterPunch nails this one

January 11, 2008

More Questions About Diebold Voting Machines
Did Hillary Really Win New Hampshire?

PS Lindorff claims that Kucinich has filed for a recount



just pixels:

People keep attributing NH results to "Bradley effect". Hardball put up the stats: polling something like Bradley 50, Dukmejian 43, undecided 7%; election Bradley 49, Dukmejian 51. In other words, Bradley moved within the margin of error and Dukmejian got the undecided. I'm not convinced race tinged polling differently than voting.

In other races I've heard mentioned along those lines, the same pattern appears. Actuals vary from late polls within margin of error and undecideds broke toward the white candidate (which may be significant).

In NH, the sample weighting didn't accurately reflect the proportion of women who turned out for Clinton. Plus polling had Obama way ahead which suppressed his turnout.



For all of the fraud pushers out there... While there is a pile of statistical evidence out there to be examined, I seriously think those pushing a fraud theory are doing so out of preference for their own candidate of choice and less out of fact. There are many, many rational reasons to the failure of the polling including ground organization, soft-support, even a change in technology (landlines vs. cell phones). Further, you add the compressed cycle between IA and NH and the 24/7 news cycle. Those who fear for the Republic shouldn't do so because of NH. They should do so because of 2000.



From my experience in some countries where polling fraud is more regular occourence it is very difficult to detect it later. Sometimes the agencies (that are usually responsible for it in many countries) even replace the actual ballot boxes and with them the votes casted to leave no evidence behind. Whether the poll/election is stolen can only be judged by how surprising the results are for the public that have participated in it.


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.