The "Clinton name recognition" meme goes both ways.

[wyldwolf]

I often hear on DU that Clinton's lead in the primaries is mainly due to her name recognition. Though I doubt that is as paramount as it is made out to be, Jonathan Singer at Mydd makes the case that "name recognition" might also be the reason Clinton performs below Obama and Edwards in general election matchups even though "unnamed Democrat" outperforms the GOP often by double digits in polling. Her underperforming numbers aren't because she is Hillary, but because she is the named Democrat most people know.

Nonetheless, it is interesting to see that, broadly, both Obama and Edwards run better in the head-to-head matchups than Clinton. This is actually borne out in other such polling as well. While I think this is largely a factor of Clinton's near universal name recognition, which any Democratic nominee would have by the time voters decide the next President in November 2008 (and thus it is not necessarily the case that Clinton would be a weaker general election candidate than either Edwards or Obama), these numbers do seem to undercut the notion shared by many Democratic primary voters that Clinton is the most "electable" candidate in the race.

http://www.mydd.com/story/2007/6/12/13021/9377

I would add that Singer's reasoning also undercuts the notion by many in the blogosphere that Clinton is the least electable of the top 3.

[Czolgosz]

1. When people already know a candidate (high name id) and don't like the candidate (high negatives) it

is very tough to overcome because the high name id indicates that the voters have largely made up their minds and, correspondingly, there is a smaller pool of voters who don't know the candidate well yet and who may still be favorably impressed as they learn more about the candidate. There is an interesting discussion of this phenomenon at the Washington Post's blog:

A generic poll by itself simply measures public attitudes toward a candidate's party, which is only one factor (but an important one) in how a voter will ultimately vote. On the other hand, at this early pre-election stage when party is unidentified and voters are asked to select among or between names, then the poll becomes much more a name identification poll.

The best reflection (at this admittedly early stage) is to poll asking something like "If the election were held today, would you likely vote for REPUBLICAN John McCain or DEMOCRAT Barack Obama"? ... The most revealing feature of polling structured in this manner is that "consensus" candidates like John Edwards who are not running at the top of the pack, but who are acceptable to a wide range of voters, will suddenly vault to the top. At the same time, well known candidates who also have high negatives (Hillary Clinton) begin to bump against their numerical ceiling.

http://blog.washingtonpost.com/thefix/2007/04/parsing_the_polls_is_the_white_1.html