Mike who??

Wait, these polls can actually change over time? Why if that were true, it would imply maybe they are NOT reliable indicators of who is the strongest nominee.

Surely Eric did not mean to admit that.

Helter,
McCain has consistently done well in national match-ups, probably better than other GOPers on average. With Romney and Giuliani getting the majority of press, the media is no longer going after McCain with the intensity they were earlier. Since he is being relatively ignored these days, McCain can once again use the "outsider" theme. I've been saying for a while that it would be either Huckabee or McCain. I don't see any reason to change that prediction. I was cheering for Tancredo, but I don't think that is going to happen unfortunately.

Helter,

The basic explanation is that for their party ID mix, SurveyUSA got 42% Democratic, 30% Republican, 24% independent. Note that would be a huge change from the 2004 election, where exit polls showed the mix as 40% Republican, 35% Democratic, 25% independent.

As is typical in these polls, Obama overall did better than Clinton among Republicans and independents. Where he lagged Clinton was among Democrats. Doing well among the Democrats plus the party mix explains why Clinton did relatively well against the other three Republicans (even though she actually lost the independents to Giuliani and Huckabee, whereas Obama won among independents against Giuliani and tied with Huckabee).

But unlike with the other Republicans, that party mix was actually relatively favorable for McCain. He won the independents by 20 points against Clinton, 56-36 (although only by 8 points against Obama, 48-40). Most importantly, he got 25% of Democrats against Clinton, more than any of the other Republicans (consistent with the overall pattern, McCain did even better among Democrats against Obama, getting 30%).

Of course those cross-over numbers are unrealistically high, and will very likely go down significantly once the nominees are decided. Still, this is a good illustration of how even in a case where party ID massively favors the Democrats, they could actually lose if their nominee loses the cross-over and independent battles badly enough.

To say that the poll shows Clinton to a be a stronger nominee than Obama is overdrawing things a bit. Ohio has not been the site of much campaigning, yet. Thus, Clinton's initial lead in visibility is still a factor. However, even now, the differences between her performance and Obama's in projected races against Republicans are small. In fact, this poll tells us next to nothing about what would happen 11 months from now, when the country votes.

mikeel,

I don't see why the Democrats can't beat McCain.

Consider for example this Iowa poll:

http://www.surveyusa.com/client/PollReport.aspx?g=2d211d2a-ee13-40ab-8455-5a70b9b55bfe

Iowa is as close to a legitimate general election test as we can get, since at least the primary campaigns have been active there for a long time. And in that poll, Obama (although not Clinton) came out ahead of McCain, in virtue of keeping it close among independents and actually winning more Republicans than McCain won Democrats.

That of course is the logical extension of the point above: McCain can win the Presidency despite the Republicans' likely party ID deficit because he has cross-over and independent appeal. But if the Democrats also run a candidate with cross-over and independent appeal, it negates that effect and their party ID advantage (39-29 in this Iowa poll) wins out.