Sanders leads Clinton in new NH poll….

In a new Franklin Pierce University/Boston Herald poll  Bernie Sanders now leads Hillary Clinton 44 to 37%….

The poll is small, 442 people and has a high margin of error, 4.7%….But it IS the first poll to show Sanders ahead of Clinton….

We’ll see what other polls say….

The Vermont Senator has been gaining steadily on Clinton in neigboring New Hampshire…

A stunning new poll has Sen. Bernie Sanders (I-Vt.) beating presumptive Democratic presidential frontrunner Hillary Clinton in New Hampshire.

Sanders has eclipsed Clinton by a 44 to 37 percent margin, according to a new Franklin Pierce University/Boston Herald poll that was first reported by the Boston newspaper Tuesday evening.

The previous FPU/Herald poll taken in March had Sanders trailing Clinton 44 to 8.

Today’s poll is the first to show Sanders, whose liberal policies are popular with the Democratic base, ahead of Clinton in New Hampshire….


Correction….The original number of people listed for the poll

BY ME was 4,442…. That is incorrect…The corrected number is 442 respondents……

Share on Facebook

17 thoughts on “Sanders leads Clinton in new NH poll….”

  1. BTW?

    Iowa Democratic Presidential Caucus
    (D)Clinton 52, Sanders 25, Biden, O’Malley 7, Webb 3, Chafee 1 Clinton +27

  2. It may not be intuitive, James, but 4,442 is a huge sample for a state the size of New Hampshire.

    I think that when I took the California Poll for Field Research Corp. in the early 1980’s, the sample for the whole Golden State was about 3,000, and if you look at the national samples at, they seem to be about the same size.

    I’m no good at calculus or statistics, but this where the margin of error and the 95% confidence level for random-sample polling kicks in. If your sample is truly random, and also balances the most important geographic and demographic criteria, then 19 out of 20 of your samples will match the entire population within the Margin of Error.

    Ideally, you might want a sample of 10% or 20% or 50% of the population (of course in a real election you want as close to 100% as possible), but the relatively-small increase in confidence and accuracy that you’d gain by multiplying the sample by 1,000 or 10,000 doesn’t merit the extra time, money, manpower and computation that would be involved.

  3. Yeah, I’m actually surprised that with that sample size, the margin of error is so high. I don’t have my old stats text available, but I assume that the survey is “correcting” for some sampling bias.

  4. New Hampshire is a giant killer some times.

    Her ads are phony log cabin stuff.

    Nasty problem those E mails too.

    Still lots of time.

  5. Ultimately, I don’t think too many people will vote against someone because she/he erased emails. I’d bet there will be bigger issues by next November.

  6. DSD I sincerely apologize
    The poll was with four hundred forty two people….
    Not 4,442…..
    It’s a small poll with a high MOE


    Sorry about that

  7. Sanders leads in the poll…

    But we’re ALL gonna be waiting for confirmation from a few OTHER polls…

  8. Politico goes into the Bernie the Socialist thing….

    “You don’t change the system from within the Democratic Party.”
    “My own feeling is that the Democratic Party is ideologically bankrupt.”
    “We have to ask ourselves, ‘Why should we work within the Democratic Party if we don’t agree with anything the Democratic Party says?’”
    Bernie Sanders, everybody—the same Bernie Sanders who is running to become the Democratic Party’s candidate for president of the United States.
    The most surprising thing about the independent Vermont senator’s surprisingly successful campaign so far is not that he’s doing it as a self-described democratic socialist. It’s that he’s seeking the nomination of a party he caucuses with in the Senate but is not a part of, isn’t a registered member of and has never been a registered member of—a party he’s spent his 40-year career beating at the polls and battering in the press.

    More @

  9. I glanced at Wednesday’s Boston Herald and it said 442 (we all know what it’s like to hit a key too long or once too often).

    That’s a reasonable number, especially since it was asking Democrats in NH, and compares with the samples taken in Rhode Island, with a similar population and the same number of Congressional seats (2).

    With over 400, you can conveniently allocate the sample to about 200 in each C.D. for district results with a higher margin of error, but one that is still within a useful range.

  10. Second Iowa Poll…..

    Iowa Democratic Presidential Caucus
    Clinton 50, Sanders 31, Biden 12, O’Malley 1, Webb 1, Chafee 0 Clinton +19

  11. The Clinton people are saying this NH poll had unaffiliated voters including in the sampling…
    Everyone is waiting for a second NH Democratic poll to come out..

  12. It all depends on how closed a party’s primary or caucus-tree is.

    If unaffiliated/unenrolled/decline-to-state voters are allowed to vote in partisan primaries, or if they’re allowed to switch to a party close to primary election day and then vote in that party’s primary, then you want to include unaffiliated voters in your sample; in fact you’d be skewing your sample and results by excluding such voters.

    If on the other hand it’s hard for unaffiliated voters to vote in partisan primaries, then it would be equally mistaken and misleading to include them in a party-primary sample.

Comments are closed.