OK, I'm sorry to bang on about the polls again. I know that in speculative discussions like this, in the end we're all down to our own, individual assessment - based on our personal observations and informed guesses. I also know that you don't think much of polls this far out from the elections, Joe, and it's certainly true that there are no data we can pluck from future times to corroborate our ideas. But there are data out there that at least
can make our guesses informed ones, or that can at least serve as a useful, preliminary double check for our assumptions.
Polls, of course, have many limitations. They cannot predict the future, not even what will happen next month. They can only say what people are thinking right now, and those thoughts are tentative, because the overwhelming majority of people has barely tuned into the race yet, and knows only a modicum of what there is to know about, for example, Rudy's past. (Though it is easy to overestimate how much the average voter will have absorbed even by election day). To some extent polls still reflect name recognition more than anything else at this time, and results fluctuate quite a bit, especially when people are not just asked to name a preferred candidate but to choose sides in hypothetical match-ups.
But they do constitute a resource that represents a range of pollsters regularly surveying a cross-section of Americans about exactly the questions we're speculating about here now. You say, Joe, that Rudy has the uniquely potential to lead the Republicans to self-destruction, while other candidates would at least keep the damage limited. You say that Hillary is uniquely constrained by widespread "Hillary hatred" to a close election result at best, while Obama and Edwards have the potential to appeal to swathes of long-held Republican constituencies. This, I assume, is based on readings of what the newspapers and magazines say, and personal observations of how friends, family, colleagues and acquaintances respond. But is any of it borne out at all by the wide range of surveys done of actual cross-sections of the American voters?
The short answer is, no - none of it is. Being the electoral geek I am, I maintain an Excel file in which I track the results of so-called "match-up polls". These are the polls where the respondent is asked, "if the candidates are, respectively, X for the Republicans and Y for the Democrats, whom would you vote?" Some polls ask only about Hillary versus all the four Republican frontrunners, or only Hillary vs Giuliani and nothing else, and some match up Obama and Edwards too, to one or some or all of the Republicans. There's fewer than I would like, but there's been many since the campaign first kicked off last spring.
Here you see the graph for the polls that matched up any or all of the Democratic frontrunners against Giuliani since late March. The zero line represents a tie: say, Hillary and Rudy both get 47%, or both get 43%, with the rest undecided. When the bars dive below that line, it means the poll gave the lead to Rudy; the higher above the line it goes, the larger the hypothetical victory of the Democrat. Blue is Hillary vs Giuliani, brown is Obama vs Giuliani, yellow is for Edwards. The trendlines represent a running average.
What do we see? For some time, back in between March and June, Hillary did have a disadvantage compared to the others. Not a huge one, mind - nothing like the kind of gaping difference of performance Joe describes as probable. If the election had been held right then, Obama and Edwards would have won against Rudy by 51% to 49%, or 51% to 48% - a two or three point margin; while Hillary would have lost by an equally tight margin. Either way it would have been the kind of election we saw in 2000 and 2004, camp against camp, shifts of an inch this way or that at most.
That picture has basically persisted throughout since. Even if elections were held today, the margin of victory would be comparably small. The only difference is that for four months now, Hillary does as well as Obama or Edwards does. There is no noticable difference whatsoever anymore between Hillary's appeal and Obama's, in a race against Rudy.
Now here's how the Democrats have been stacking up against McCain:
The picture is remarkably similar. Margins are small, much smaller than you'd expect if you go by polls asking people about their opinion of Republicans and Democrats generically. McCain, like Rudy, does better than the Republican brand does overall.
Nevertheless, McCain, too, would stand to lose a general election against any of the Democratic frontrunners if elections were held today, or if they'd been held any time in the last few months. Polls are sparser than for Giuliani so the estimate is more tentative, but he'd currently lose by a four- or five-point margin; i.e., Hillary 52%, McCain 47%. Or Obama 52%, McCain 48%.
The other parallel is that again, Obama and Edwards were doing somewhat better than Hillary at some point in time; they'd lead by a 5-6 point margin instead of a 1-2 point one. (That's Obama 53/McCain 47 instead of Hillary 51/McCain 49; still hardly a collapse of the Republican coalition overall in either case.) But here, too, that difference seems to have disappeared altogether. In the last five polls that matched up both Hillary and Obama against McCain, Hillary actually fared
better than Obama four out of five times.
Next: the Democrats versus Fred Thompson:
Fred's name only started popping up later in the race, hence the lack of polls early on; and he's been fairly discounted for a while now by the pundits, so there aren't a lot of polls for the last month and a half either. But all in all there's still been a fair number.
The first thing to notice here, re: the assumptions made in this thread, is that Fred does a LOT worse than Giuliani in these match-ups. On average, we're talking a 10-point lead for the Democrat here: the equivalent of a 55% vs. 45% election outcome. Now that would be a landmark election result, though it bears pointing out that part of the reason is that there is a larger number of people who just dont know enough about Thompson to make a choice for him, so in reality it's more like 50% vs. 40%. Still, it's clear that, at this time at least, Fred's "floor" of support is far lower than Rudy's.
Then there is how Hillary appeal stacks up in a race against Thompson compared with Obama's or Edwards'. There is no trendline for Edwards in these graphs because there are not enough polls with him, but his results tend to go fairly toe-to-toe with Obama's. What we see is that Obama did have the full-scale breakthrough numbers initially - a 15 point lead! But as the campaign progressed and people found out more about both him and Fred Thompson, his lead diminished somewhat - while Hillary's increased. For two months now, it wouldnt have mattered whether Obama or Clinton had stood if elections against Thompson as Republican nominee had been held right then - they'd perform equally well.
Then there's Mitt Romney - the candidate Joe thinks is most likely to win the Republican nomination -- and I'm thinking (and wishing) he might well be right. Here's the graph:
Dont be fooled by the size of the bars here: I've adapted the value axis especially for Romney, simply because of the huge defeats he faced in these polls against the Democrats. So where the first horizontal line above zero represented a 5% lead for the Democrat in the other graphs, here it's 10%. Back in May, Edwards and Obama were beating Romney by margins in between 15% and 35% - you're talking the equivalent of "Barack Edwards" 62%, Romney 37% here.
That, of course, again was largely also a question of lacking name recognition for Mitt - in reality, it was more like Edwards 55%, Romney 30%. Striking numbers nevertheless. The race has tightened since - but still only to margins that are actually worse than Fred Thompson's. Even now, Obama has a ten-point virtual lead over Romney; Hillary a 12-point one; and Edwards a close to 15-point one.
Romney's numbers, like Thompson's, seem to reflect the strong generic preference for a Democrat over a Republican at this time. Only a strong candidate can push the Republicans into a much closer race. Taking into account the chance that they could still self-destruct as the campaign wears on, Rudy and McCain at least show that they have the
potential to do so. Romney's numbers do not offer any such signs so far, despite a campaign that's been more prolific than either Thompson's or McCain's.
Finally, notice that this match-up, too, has seen Hillary doing as well as Obama and Edwards for a while now. In fact, her numbers have remained roughly stable (somewhere between +5 and +10), while Obama's started higher but have dropped to the same levels since.
This raises the question of what could happen when the full-scale campaign starts. Hillary at least has the advantage of being a known factor: it's hard to imagine what her Republican contenders could dig up about her, or play up about her, that would still change impressions of her drastically. Obama, on the other hand, has not been attacked in anything like the same way yet.
We have no idea what impact full-throttle attacks by the conservative machine would have on his numbers if he is elected nominee and they do come. Perhaps they'll slide off of him like water off a duck: some people have that teflon-like charisma. What we do know is that in many of these graphs, we see that even in the course of a half year of campaigning in which he received
overwhelmingly positive media coverage, his initially high margins over time dropped to a level that's comparable to Hillary's.
In short - for what it's worth, and with all the caveats that come with opinion polls in general and ones this far ahead of elections in particular - there is nothing in the data that we do have about the respective candidates' appeal to the electorate that would confirm the assumption that Obama or Edwards would far outdo Clinton, or that Giuliani is the most likely to bankrupt the Republicans' numbers. If anything, they suggest the opposite.
Didn't he just sit around on his thumbs like a pantyweight in 2004? I'm having a little trouble believing this is true...