I don't know. I think there may be a third view. And that is as one approaches the election day, the vote tends to narrow because (a) people who have been on the sidelines start to get interested, and (b) there is a general sympathy for whoever is the underdog, regardless of incumbancy, and the momentum goes his/her way in the waning days.

I think that much of the "conventional wisdom" and many of the "cardinal rules" of poll interpretation are probably very shaky, particularly in an era when a large number of people have caller ID and may simply not answer the phone when the caller is unknown.

My purely anecdotal observation is that quite often the "late polls" during the election season are quite disparate from the actual outcome . . . not necessarily in identifying a winner, but in identifying percentage breakdown.

I would probably avoid making too many direct comparisons between the CT results and Gov. Carcieri's race or that of the outgoing incumbent junior senator, for the simple reason that the former is involved in a primary, and the latter is not.

Primary voters simply act and vote differently than general election voters. As you know, primary voters tend to be more issue-driven or "ideological," whereas general election voters, while possibly still somewhat ideological, are usually more civic-minded or duty-driven. Primary voters are the ones who make that little extra effort to get out and vote, because they believe their vote will "count more," for lack of a better term. And generally speaking, they are right.

Despite most MSM polls showing Chafee "ahead" of Laffey (not directly, but supposedly, as against Whitehouse in the general election), though both still supposedly behind Whitehouse in a general election matchup, I simply do not believe them. What people actually DO voting in a GOP primary, is often quite different than what those same people say they will do later on. The "pro-Chafee" and "pro-Laffey" voters will slant a pre-primary "general election" poll, so as not to favor either of them (against the Dem), because they will often give answers in polling, as if they would not vote for the other GOP primary person in a general election, when pitted against the Democrat nominee. While there are people like that (like me, perhaps), they are in the minority.

In addition, the Democrats actually end up slanting polls in their favor, because at this point, they basically have one choice. They're being asked if they'll vote for their "only choice." If they believe the choice has already been made, it is impossible for them to motivate 100% of those folks to actually translate what they have said, into how they'll actually vote. It simply never happens.

Most people who vote in their party primary, also end up voting in the general election, and generally, for the member of their own party. When given a choice, most people choose between the choices they have in front of them (of course, unless you're Linc, then you write in someone) OR they stay home. Primaries tend to get people who want to send a message with their vote out of their homes (good for Laffey). I'm convinced either Chafee or Laffey, can still beat Whitehouse (of course, given a choice, I'm still most definitely with Laffey). It will still be tough for either of them, but it can be done.

It's hard to see the same trend benefitting both Carcieri and Chafee. Hardcore conservatives love The Don and hate Chafee - Carcieri only endorsed him because he was shamed into it by some national party people.

And here I thought "hardcore" conservatives only hated pollsters (merely loathing Chafee). That's how I interpreted Kerry's laughable exit polling pratfall.

I've often wondered if exit polling has any value, from the "moral values" canard thrown around after the '04 election to the whole question of whether voters are honest about who they voted for.
Does anybody else think that voters lie to build up the hopes of the other camp when exit polls are shown, only to make the fall that much more painful and shocking?