About this Blog

RealClearPolitics HorseRaceBlog

By Jay Cost

HorseRaceBlog Home Page --> January 2008

The Giuliani Campaign, RIP

It is easy to criticize the Giuliani campaign, but it is also easy to overdo it. Rudy did not play the hand he was dealt very well, but the fact is that he was not dealt a terribly strong hand.

First, the criticism. Late last year, the Giuliani campaign pinned its hopes on no single candidate emerging from the early contests as a clear frontrunner. It got exactly what it wanted. Huckabee checked Romney in Iowa. McCain checked Romney in New Hampshire. Romney returned the favor in Michigan and Nevada. McCain checked Huckabee in South Carolina. Five states. Three winners. This was basically what Rudy wanted (though he might have preferred a Huckabee win in South Carolina).

So, what happened?

The Giuliani campaign correctly figured that this scenario would not give a single candidate overwhelming momentum. This in turn meant that there was not a prohibitive favorite going into Florida. However, Rudy's operation ostensibly failed to understand that momentum works in the other direction, too. Viability, likability, and vote choice move together - in both ways. Voters flock to winners, and they flee losers. Loss after loss after loss - and Rudy started to look like a loser. As he seemed less viable, his numbers nationwide (and in his "firewall" of Florida) dropped.

This account is as relevant as the media's (somewhat self-serving) explanation that his numbers dropped because he was not in the news. While there is much truth to this, it fails to account for the fact that, insofar as he was in the news, it was negative coverage of him withdrawing from race after race. There were also many opinion pieces that questioned the wisdom of this strategy. Stories like these damage a candidate's viability. The Giuliani campaign thus experienced the awful process of negative momentum that Larry Bartels outlines. Voters come to see the candidate as a loser, they like him less, and they start to abandon his candidacy, thus making him look like a loser all the more.

The Giuliani campaign had other problems. Bernard Kerik was indicted on November 8, 2007 - and an examination of the GOP national chart shows that shortly after this event Giulaini's numbers began to fall. In fact, if you look at his trend line, you see two distinct periods when his numbers dropped. The first is this post-Kerik fall in November. He fell from the high twenties to the low twenties. The second occurred in January, after the start of the early contests, when he fell from the low twenties to the tweens (presumably) because of the negative momentum. So, I think it is reasonable to argue that Kerik's legal troubles contributed to Giuliani's political troubles.

Speaking impressionistically, my sense is that the Giuliani campaign had message difficulties, too. Good campaigns begin with good biographies. Rudy has an impressive professional biography. However, good campaigns use these biographies to develop coherent, compelling messages. For the best campaigns - there is a point at which biography and message become indistinguishable. We saw this with Clinton-Gore '92. We are also seeing something like this with Obama. I do not think the Giuliani campaign ever approached this level. It never had a message that emanated forcefully from its candidate's biography. It surely tried to create one. That is what its "Twelve Commitments" were all about. So also was its rather silly line, "I don't hope for miracles. I expect them." But none of these attempts caught on.

Rudy probably had to do more work with message than other candidates. Consider that nobody has ever won the presidency having most recently been mayor of a city. This is important. Forty years ago, Joseph Schlesinger argued that there is an "opportunity structure" to American political careers. In Ambition and Politics he asserts, "American political careers do not proceed chaotically. There is a pattern of movement from office to office." Holding a given office yields opportunities for promotions to some higher offices, but not others. To date, a mayoralty has never yielded the presidency. So, in a sense, Giuliani was jumping the line.

Did this have an effect? Perhaps. Like everything in electoral politics - the structure of opportunities influences and is influenced by voter perceptions. I know a few voters who looked at Giuliani and said, "He was mayor of New York. How does that qualify him to be president?" I am not sure if this was a widely held opinion, but I reckon that it might have been. And if it was - it was a problem. I think he could have mitigated it with a more compelling message, one that vividly connected his success in New York to his agenda for Washington. He never developed this. His attempts to do so seemed rote, uninspired, and based on the assumption that voters would make the connection themselves. So, perhaps it is unsurprising that more Florida voters saw McCain and Romney as better prepared to be commander-in-chief.

That's enough picking on Rudy for me. Like I said, it is easy to take this too far. For Giuliani had a real disadvantage that most (including myself) failed to perceive until afterwards. Simply put, the early states were on friendly territory for Huckabee, McCain, and Romney - but not really for Giuliani. The three winners could make niche appeals that Rudy could not. Huckabee has a connection with evangelicals, who made up a large chunk of the Iowa electorate. Romney and McCain have connections with New Hampshire. Romney grew up in Michigan, and he could rely on strong Mormon turnout in Nevada. As Tom Bevan has pointed out, each victor won his respective state by running as a candidate for "governor." At best, Rudy had an angle in New Hampshire - but it was a relatively slight one, given McCain and Romney's ties to the state.

Also, we must remember that Giuliani's Florida firewall plan was a strategy borne of necessity. Contrary to the conventional wisdom, Rudy competed in Iowa early on. He made 43 appearances there. He spent money there. He pulled out in August - when it became clear he could not win. That goes double for New Hampshire, where he held 87 events (almost as many as he held in Florida) and spent loads of cash on advertisements. He pulled the plug when it appeared he could not win, and perhaps did not want to impede McCain from beating the well-heeled Romney. So, it is simply wrong to suggest that his Florida strategy was his first choice. It was, in fact, a fall back.

All in all, Giuliani was not in nearly as strong a position as last year's polls implied. Instead, he had some real disadvantages vis-à-vis the early schedule. Unfortunately, his campaign did very little to help itself. It wrongly thought it could abstain from the early states without suffering a penalty. It never found a way to handle the Kerik indictment. It never developed a compelling message - something of great importance for a candidate with Giuliani's particular biography.

-Jay Cost

How McCain Won

John McCain won Florida by putting together the same basic voting coalition he forged in New Hampshire and South Carolina. What is impressive is that he did it in a closed primary. Registered Independents and Democrats were not allowed to vote, but McCain still won. Let's take a look at how he did it:

-McCain once again won those who are disenchanted by the Bush presidency. Most Florida Republicans (68%) approve of the Bush administration. Romney won them, 35% to 31%. McCain, however, scored an overwhelming, 22-point victory among the 32% of voters who disapprove. I think this is one of the evolving stories of the Republican contest. If you like Bush, you are inclined to Romney (or one of the other candidates, all of whom but Ron Paul do better among Bush supporters than Bush opponents). If you dislike Bush, you are inclined to McCain.

-From a certain perspective, this is an ironic feature of this campaign. McCain has been campaigning, in part, on the surge - the hallmark of the Bush presidency for the last year. Romney has been campaigning on fixing Washington. But the results do not follow these pitches. Why? I think one reason has to do with the long memories of voters. McCain's reputation as an anti-Bush maverick is still quite ingrained in their minds. So, those who disapprove of Bush are "naturally" inclined to McCain, despite Romney's anti-Washington pitch. Meanwhile, voters supportive of Bush recall how many times McCain has been a thorn in the president's side, and so are inclined to Romney.

-There is a lesson in all of this about the limitations of political campaigns. They only do so much to shape the thinking of the American voter. Those who have held opinions about political figures for a long time are not going to be easily disabused of them, despite how many political ads are run or adjustments in messaging are made. I think this hints at a mistake the Romney campaign made - it pivoted too late to a message about fixing Washington.

-McCain won voters for whom the economy is their top concern, 40% to 32%. Remember that McCain won them by a nose in New Hampshire. Isn't that strange, given Romney's message? Not necessarily. If we step back and look at it from a broader perspective - this can start to make sense. While it is true that Romney's campaign message has been about fixing the economy - Romney won voters who think the economy is healthy. McCain won voters who think the economy is sick. So, it should come as no surprise that the voters for whom the economy is tops went for McCain, given these divisions. If you think the economy is healthy, it is probably not your top concern.

-McCain won the ideological coalition he won in the previous states. He won liberals and moderates by a large margin. He split those who consider themselves "somewhat" conservative. And he lost those voters who consider themselves "very" conservative by a wide margin. We saw this in New Hampshire and South Carolina.

-My sense is that there is some causal connection between evaluations of Bush, evaluations of the economy, and ideological orientation. They influence one another - though we cannot say which opinons are primary and which are secondary (or if there is something else entirely that is driving all three). And, of course, we have to acknowledge that there are exceptions (e.g. the good number of Bush supporters who went for McCain). But I think that these three characteristics are somehow related to one another - and so many voters fall on one side or the other. McCain, as the candidate with a long history of being the anti-Bush "maverick," has an advantage with voters on one side of the divide, despite Romney's electoral pitch. Unfortunately, I can do little more than theorize about this relationship, though I think it is inherently reasonable. If I had more comprehensive exit poll data, it would be pretty easy to test whether these opinions are indeed related.

- Of course, elections can never be reduced to single causal narratives. There are several factors that do not fit into this story, but were nevertheless important.

First, like any winning candidate, McCain did well among many of the factions he lost, including Bush supporters. He lost them by just 4% last night. It was his 22-point victory among those who dislike Bush that is the noteworthy result.

Second, McCain was perceived by more Floridians as the most electable, edging Romney out by 13 points on that quality. As I have said time and again on this blog, there is a strong connection between perceptions of electability and vote choice.

Third, Romney won voters who said that cutting taxes was the higher priority, 35% to 29%. McCain won those who said reducing the deficit was more important, 42% to 27%. This, I think, shows the potential of a campaign - Romney has definitely developed some tax cutting bona fides with the Republican electorate.

Fourth, McCain won a decisive victory on the question of who is most qualified to be commander-in-chief, beating Romney by 18 points. He beat Giuliani by 30 points, which is unbelievable considering Rudy's "Test. Ready. Now." slogan. Relatedly, he beat Romney among military vets by 7 points.


Like I said, these cannot be brought under the narrative I developed above - though I think that each of them was important.

-What about the geographical distribution of the vote? Florida has four large metropolitan areas: Jacksonville, Orlando, Tampa, and Miami-Ft. Lauderdale. Romney won decisively in Jacksonville, 42% to 29%; he won a slight victory in Orlando, 33% to 32%. McCain won Tampa, 37% to 30%; he won big in Miami-Ft. Lauderdale, 45% to 22%. Unfortunately for Romney, Miami-Ft. Lauderdale/Tampa out-voted Orlando/Jacksonville by better than 2:1.

-A final point on the exit polls - bad news for Mike Huckabee. He won Iowa on the back of self-identified evangelicals. However, he only split them with Romney and McCain this time around. What is more, he garnered a measly 4% among those who do not identify themselves that way. He goes on to Super Tuesday, and I think he could be a factor in the South. But clearly his voting coalition is shrinking, not expanding.

-Jay Cost

How To Read the Early Results

The following is an email I received last week from a Romney supporter:

How do you call McCain the front runner when Mitt Romney has more delegates and has won more republican votes in all the primaries?

That's a good question, one that several emailers have asked me. To be fair, I have been very cautious in my approach to McCain's candidacy. I think he faces some serious opposition in the Republican Party. Unlike previous nominees, McCain is not a consensus candidate. Instead, he has inspired a faction of the party to oppose him. And in a system like ours, factions are often able to prevent things from happening. So, they might be able to stop him.

Beyond this, I think the emailer asks a fair question. After all, this is a race to win a party nomination. This means it is about delegates. So, shouldn't they be our focus? Aren't we wrong to focus so relentlessly on who has won which states? Doesn't that obscure the true dynamics of this race?

Well - yes and no. I think that the proper way to look at the race is somewhere in between. Yes, the delegate count is important. But who is winning and who is losing is important, too. What we need to do is find a mode of analysis that integrates each into a coherent scheme.

But first we should define our terms. We are contrasting two different ways to look at the nomination process. The first is the competition for delegates. This is what Giuliani was stressing prior to his December collapse. His campaign was saying something like, "It's all well and good who wins Iowa, but the fact of the matter is that Iowa's delegates to the convention will not be chosen until the summer." This was similar to the Romney camp's basic line after Michigan - "We are focusing on Nevada more than South Carolina because Nevada has more delegates."

The second is the "beauty contest." Up until the 1970s, that is all that primaries were good for. They did not factor into delegate allocations. They were used simply to get a read on the preferences of the party rank-and-file. So, those who view these early elections as beauty contests care only about who wins and who loses. At its extreme, this form takes primary elections like general elections. The winner is the candidate who wins the most votes. The rest are losers.

So, which of these is the better way to look at these early contests? I think both of them are insufficient. Both miss a big part of what is happening.

On the one hand, it is too hasty to start looking at delegate counts. Doing this is like identifying a vote leader after 6% of precincts have reported. Right now, Romney has a 19-delegate lead. However, there are 2,380 delegates who will decide the Republican nomination. So, Romney has 2.48% of the total delegates. Huckabee has 1.6%. McCain has 1.35%. This is a negligible lead - especially in light of the fact that Super Tuesday will allocate about 39% of all delegates.

What is more, following the delegate count is simply not the way the press looks at these races. And, regardless of how narrow its view of things is, the fact is that the press' perspective matters. When you get right down to it, many voters are essentially ambivalent when it comes to the candidates. They can be swayed by positive or negative media coverage, which almost always follows a win or a loss. We call this momentum, and its presence can make a state more important than the delegates it offers. New Hampshire is a great example. The Granite State has just 12 delegates at this year's GOP convention - but the way the results were covered launched John McCain to the front of the national polls. We can debate whether or not it should be this way; but the fact remains that it is this way.

On the other hand, looking at these early states just as beauty contests - with one candidate deemed the winner and the rest deemed losers - brings a different set of problems. Focusing exclusively on who won a plurality of votes obscures two important features. First, it matters how a candidate wins the votes he wins. What was the voting coalition that he put together? This is an important feature to know before we generalize from the state in question to the nationwide electorate, which is what is really important. Second, it also matters who he ran against. It might be the case that Candidate A beat Candidates B and C - but by the time the next state rolls around, C will have exited the race, giving B an advantage over A.

These considerations point to how we should examine the early contests. We alter the metaphor a little bit: we take the early states not as beauty contests, but as exhibition games. We look at the voting coalitions the candidates are forming in the early contests, and infer how they will factor when delegates are awarded en masse on Super Tuesday. This means that it matters who wins a plurality of votes, but this is surely not all that matters. Candidates can lose and still show signs that they are putting together a voting coalition that will deliver the nomination. This also means that we take the early contests seriously, but we do not make them out to be more than they are. When they are properly understood, they are our best indications of how candidates will do on Super Tuesday. No more, no less.

On the Republican side - this helps answer the emailer's question. As of today, Romney simply cannot be seen as the GOP frontrunner. Though this might change tonight, John McCain is currently the candidate who has put together the larger voting coalitions in more competitive states. Romney essentially bowed out of South Carolina - where three other Republicans were competing. He won Michigan, but it appears that his victory was based in part on his deep roots with the Wolverine state. On the other hand, McCain is at best a tenuous frontrunner. His voting coalitions have sampled heavily from Independents. Without them, it is questionable whether he would have won New Hampshire or South Carolina. In many Super Tuesday states, he will not be able to rely on them. And, while he has great appeal with moderates and some appeal with conservatives - he has failed to win a plurality of strong conservatives in any state. Whatever happens today in Florida, the results will probably be too close to be determinative. So, the five states to date indicate that no candidate has yet won a sustained voting coalition that could carry him to victory on Super Tuesday.

-Jay Cost

In Hot 'lanta after Super Tuesday

For those of you who live in and around Atlanta, I thought you might find this notable. Joseph Knippenberg of Oglethorpe University has invited me to participate in a conference on February 6th. It looks to be quite a fun evening with a very august group of thinkers. I'm looking forward to it. Maybe I'll see you there!

Here are the details:

On Wednesday next, the day after Super Tuesday, I've put together a little conference to pick through the entrails of the isms (liberal and conservative), the parties (Democrats and Republicans), and the nominating processes. To be held in Lupton Auditorium on the Oglethorpe University campus, "The Future of Liberalism and Conservatism During and After 2008" will be keynoted by our friend Jonah Goldberg, who will be speaking at 7 p.m.

Festivities will begin at 11 a.m. with a student panel, featuring bright lights from Berry College, Mercer University, and Oglethorpe. After a lunch break, we'll reconvene at 1 p.m. to discuss the isms. This roundtable will be chaired by the distinguished Peter Lawler, and will feature Jay Bookman of the Atlanta Journal-Constitution, Matthew Franck, Bryan McGraw (currently of Emory, soon to be of Wheaton College), and Susan McWilliams of Pomona College.

At 3 p.m., we'll discuss the nominating process led by a roundtable chaired by Berry College's Eric Sands. We'll hear from Alan Abramowitz, who writes here, our friend Jon Schaff, and Jay Cost. After dinner, we'll hear from Jonah at 7 p.m. After Jonah, who knows? There's a nice pub down the street, but drinks, unfortunately, aren't on me.

Our big sponsors are the Intercollegiate Studies Institute and several offices at Oglethorpe, with additional support from Berry College.

-Jay Cost

Demography and the Democratic Race

There is a simple explanation for Obama's victory last night: he won African American voters. They constituted 53% of the vote, and 80% of them went for Obama.

This is an incredible result. Of this there is no doubt. But it invites a question - can Obama win white voters?

Since he lost Nevada, pundits have been suggesting that he cannot. But I think the picture is much more complicated than it first appears.

First of all, we have to acknowledge that Clinton is currently beating Obama in the polls among white voters nationwide. The latest LA Times/Bloomberg poll tells the story pretty succinctly. Clinton leads Obama among white voters by 19 points. Compare this to Clinton's 9 point lead among all self-identified Democratic primary voters - and it should be clear Clinton is doing better with white Democrats than with the Democratic electorate at large.

So, the more precise question is: can these numbers change, or has Obama "maxed out" among white voters?

Let's take a look at recent results to see if we can get an answer.

Obama won Iowa by a decisive margin. And, as Iowa is 93% white, it should come as no surprise that he won white voters by 6 points. Clinton won white women in New Hampshire; Obama won white men; and, as more women voted than men, Clinton won the state. But both of these states indicate that Obama can win sufficient numbers of white voters.

But wait a minute - what about Nevada? Clinton beat Obama by 18 points among whites. She even won white men by 6 points. Doesn't that indicate that the dynamic of the race has changed, that Obama has perhaps been marginalized as the "black candidate?"

Well...not so fast.

Clinton won white voters statewide. Of this there is no doubt. But we might ask which white voters did she win? The Nevada entrance poll clearly indicates that she won white voters in and around Las Vegas - but that she and Obama at least split them outside Vegas:

NV Region.jpg

These results are really intriguing. Consider that, according to the Census Bureau's 2006 population estimate, Clark County (where Vegas is) is home to about 93% of the state's African American population. So, if Obama won Reno and split the rural counties, he must have been strong among white voters outside Vegas. There are simply not enough African American voters outside Clark County for Obama to do so well having lost white voters.

Let's expand on this. Unfortunately, the Nevada Democratic Party does not release raw vote totals. But we can look at delegates, which are a sufficiently unbiased estimate of how the votes went in each county. The following table reviews counties in Nevada according to: the percentage difference in delegates between Obama and Clinton, the percentage of whites, and the estimated population from 2006.

NV County.jpg

Obviously we cannot know the exact votes per county. These are delegate totals, not raw vote totals. Above all, we do not have breakdowns of vote by race per county. Nevertheless, this should make it clear that the entrance poll was certainly on to something. Obama won Reno (Washoe County), which is largely white. He won Carson City, which is largely white. And he did very well in the rural counties, which are also white.

It should be clear from this that it is insufficient to say that Clinton won "the white vote" in Nevada. It is better to say that she won a certain type (or types) of white voter. But what type? Why did white voters in Vegas break for Clinton so heavily while voters outside Vegas did not? Obviously, the ideal explanation is one that accounts for not just Nevada, but also Iowa and New Hampshire. I see three hypotheses that could connect these dots:

(a) It is a matter of GOTV organization. Obama beat her in Iowa. Clinton beat him in New Hampshire and Las Vegas.

(b) It is a matter of income. Whites who make more money tend to support Obama. Whites who make less money tend to support Clinton.

(c) White voters in racially uniform areas are more attracted to Obama that white voters in racially diverse areas.

Any of these could be true. Each of them has evidence to support them, and none of them excludes any other. I am sure that there are other potential explanations as well. Unfortunately, we cannot arbitrate between them. We could if the media chose to release raw exit data numbers, or at least more detailed cross-tabs. But they don't, so we can't. [The biggest difficulty is with the second explanation. We clearly saw income play a role in Iowa and New Hampshire - but we would need to see data on income controlling for race in Nevada, which is more racially diverse than Iowa or New Hampshire. The media does not provide that kind of data.]

The situation in South Carolina is not nearly as mixed as it was in Nevada. The fact of the matter is that Clinton won a strong plurality of white voters. Once again, gender was a critical factor. The entrance poll shows that white women broke decisively for her (42% to her, 36% to Edwards, 22% to Obama). Edwards won the white male vote, and Clinton and Obama were in a statistical tie for second. So, Obama's victory was dependent upon black voters. We can confirm this with a look at the county-by-county results. I ran an ordinary least squares (OLS) regression analysis on the South Carolina results. This is a way to predict a dependent variable based upon an independent variable. I tested whether the percentage of white residents per county could explain the percentage spread between Obama and Clinton per county. It does. In fact, it explains about 70% of the difference between Obama's share of the vote and Clinton's share. As the white population in a given county increases, Obama's margin of victory over Clinton decreases.

This lopsided result among white voters is consistent with the second and third hypotheses I listed. It could be a matter of economics. South Carolina as a whole tends to be less wealthy than Nevada, Iowa or New Hampshire. It could also be an issue of racial mixing. Counties in South Carolina are much more heterogeneous than the other three states. But that does not mean that either theory is true. All it means is that they are consistent with the data we have. These two theories help us link the results of Iowa, New Hampshire, Nevada, and South Carolina. But I am sure there are other plausible theories that could do the job.

The bottom line - this is a real puzzle, and it does not admit of any easy answers. White voters in Iowa and New Hampshire embraced Obama. White voters in South Carolina did not (though he tied Clinton among white men). And in Nevada, what they did depended on where they were located. Minimally, it should be clear that Obama can win white voters - so, the answer to our initial question is yes. Accordingly, breaking this race down by a simple dichotomy of white and black is oversimple and empirically unsatisfactory. There is more going on here than what appears at first glance. What it is, we cannot know.

This is what we do know. Clinton has done well among Hispanics. Obama has done well among African Americans. Depending on where and when, white voters vary their support. How will that play out on Super Tuesday? We can get a sense from the following table, which reviews the states on Super Tuesday, their pledged delegates, and the percentage of their residents who are white, African American, and Hispanic:

Super Tuesday Demography.jpg

The final row is worth taking careful note of. This is an average of each demographic group weighted by the delegates per state. As you can see, white voters make up a majority of the Super Tuesday population - but African Americans and Hispanics are important minorities. And remember that in most states the electorate will probably oversample from African Americans and Hispanics, who tend to vote Democratic more than whites.

And so, it seems to me that Super Tuesday depends upon three variables:

First, will white voters follow the pattern they followed in Iowa and New Hampshire, in Nevada, or in South Carolina? And remember that viability, favorability, and vote choice can go go hand-in-hand in the primaries. The current polling that shows Clinton with a large lead among whites could change as a consequence of Obama's South Carolina win. After Obama's victory in Iowa, Clinton's margin among white voters shrunk to just 8 points, according to ABC News/WaPo.

Second, will African Americans "out perform" Hispanics? This is an interesting question. The Nevada entrace poll found that African Americans and Hispanics each comprised 15% of the total electorate. But Nevada as a whole is 6.6% African American, and 19.7% Hispanic. If Super Tuesday African Americans "out perform" Hispanics as they did in Nevada - then Obama will be in a better position than what the above table suggests.

Third, what happens to Edwards' voters? It is unclear what they will do if Edwards drops out. It is also unclear whether Edwards can sustain his current support. Voters can be brutal with their evaluations of viability. If they see Edwards' candidacy as hopeless, it is quite possible they will abandon it even if he stays in. If they do, where do they go?

-Jay Cost

What's Happening in the GOP Race?

We're just five days from the Florida primary - the last "beauty contest" before Super Tuesday. Here are my thoughts on where the race stands at the moment.

(1) While the race in Florida is tight - it appears today that Giuliani is not in the lead. So, a question worth asking: if he can't win, whom does he hurt? Survey USA's recent poll of Florida Republicans is helpful for answering this question. [Bear in mind that SUSA found McCain with a larger lead than other pollsters, and Romney at a lower point than the others. Also, see the footnote at the end of the post.] Unsurprisingly, it found that Giuliani was cutting into one of McCain's strengths: moderates and liberals. However, it also found Rudy in a tie with Romney among self-identified conservatives. Even if Romney's strength is underestimated in this poll - it nevertheless remains that Rudy is drawing a good chunk of conservatives into his camp. So, it is unclear who, Romney or McCain, is at a greater disadvantage because of Rudy's presence in the race.

(2) SUSA's cross-tabs on voters' top concerns are also worth noting. The top three issues, in order, are the economy, terrorism, and immigration. Giuliani has the edge on terrorism. No big surprise there. Romney has the edge on immigration. Again, no surprise. But McCain has an edge on the economy. And remember that McCain won voters whose top concern was the economy in New Hampshire and tied them with Mike Huckabee in South Carolina. Romney won them in Michigan, and the recent LA Times/Bloomberg national poll showed Romney with a statistically significant, six point lead over McCain as to who is best equipped to deal with the economy. Interestingly, preferences on this question break along self-identified ideology - moderates/liberals and independents split between Romney and McCain, conservatives go for Romney; furthermore, the poll found that moderates/liberals and independents were more concerned about the economy than conservatives. All in all, I think the data tells a mixed tale about the political effects of the economy. I think it is fair to say that Romney's background gives him real potential with voters concerned about the economy - but he has not yet broken out with them in the early states beyond Michigan.

(3) I find it hard to make an argument about where Fred's voters will go. I thought the low-profile nature of his withdrawal was quite intriguing. Perhaps unsurprisingly, Insider Advantage's latest Florida poll, taken after Fred dropped out, shows 7% supporting "Other." Rasmussen has that figure at 6%. Any guesses as to who this "other" candidate is? It seems to me that at least some portion of Fred's voters will not learn that he is out until after they vote for him. Assuming the Fred Heads do learn about his departure - it is hard to guess whether any candidate will enjoy a significant gain. Most of Fred's voters are conservative - and as I indicated above, there seems to be a rough tie among conservative voters, with Giuliani and Romney out front by a few ticks over McCain and Huckabee. My guess is that the tie among conservatives between Giuliani and Romney is in part a function of the particulars of the SUSA poll - so I think that Romney stands to gain the most. But Fred's voters will probably spread themselves out among the rest of the field - and between this and those who stick with Fred, I am not sure that Fred's departure will have a big effect in Florida.

(4) As of now, the Florida race is as tight as any we have seen this cycle. It might be tight to the end, or it might break in one direction or the other over the weekend. The RCP average shows the candidates separated by just a few points. From poll-to-poll you see candidates trading different positions (although none at this point show Rudy in first, and only one has him in second). And, of course, 8 to 10% of the public still claims to be undecided (not including the lingering Fred Heads).

(5) I am not convinced that Florida will alter the dynamics of the race in a significant way. Of course, it might. New Hampshire - as it often does - catapulted its winner to a national lead. And Michigan seems to have given Romney a boost in Florida (and maybe nationwide, too). Florida might have a similar effect. But there are reasons to think that Florida's effect will be modest. Consider:

(a) Races like this do not end with a multi-candidate scrum - with the winner of the state in question claiming the nomination, and the opposition all falling off. They are better characterized as a war of attrition - with the field being whittled down bit-by-bit. Florida might help to whittle the field down, but we should see more than two candidates walk away from the race with a claim to viability.

(b) What is more, Florida might not whittle at all! McCain, Romney, and Huckabee have either the resources or the poll standing (or both) to stay and fight through Super Tuesday. So, the only candidate who could be whittled away is Giuliani. And, if he wins, he's still in.

(6) So, looking ahead to Super Tuesday - what do we see? One noteworthy feature is the large chunk of Republican delegates coming from blue states. Blue states don't out-number red states on that day, but they are delegate rich: California, Connecticut, Delaware, Illinois, Massachusetts, New Jersey, and New York combine to offer up 465 bound delegates. That amounts to 36% of the total number of delegates that a candidate needs to win. The delegates in all of these states except Illinois and Massachusetts are chosen via some kind of winner-take-all procedure. By contrast, twelve red states will offer up 456 bound delegates, and five of these states have some form of proportional allocation to dole them out. A strong show in the blue states was the linchpin of Giuliani's strategy - his plan was to stay viable through Super Tuesday and perform well in the blue states. Can McCain do the same if Rudy falters? He could. And, as a feather in his cap, Arizona offers up 50 bound delegates that day, winner-take-all. All in all, this could produce an interesting dynamic on Super Tuesday. What happens if the moderate McCain wins the blue states, and the conservative Romney wins the red states?

* - Historically speaking, I have not been the biggest fan of SUSA. Upon reflection, I do think it is to their credit that they offer so many cross-tabulations. That's a sign that they are not afraid of people taking a close look at their numbers. Nevertheless, it is less than ideal to have just SUSA for cross-tabulations. But to my never ending frustration, few other pollsters are so generous with their data. This makes sophisticated analysis of these races very difficult. I used SUSA to draw these inferences because theirs is the only data set available. This amounts to a real caveat. Just like the drunk who looks for his keys under the streetlight - I am dealing with the data that is available, not the data I would love to have.

-Jay Cost

Fred '08, RIP

The poor guy never had a chance.

He was caught in the Catch-22 that has bedeviled so many presidential candidates in the modern era. He was right on all the issues - but he could not get anybody to vote for him.

They won't vote for you because they don't see you as viable, his advisers told him. So, he asked, how do I appear viable? Simple, they explained, get people to vote for you.

I am speaking, of course, about Duncan Hunter.

Fred Thompson's candidacy is a completely different matter. Fred landed in the same Catch-22 that plagued Hunter (and Biden and Dodd and Richardson) - but he was the only one of the bunch who put himself in that spot.

Fred had two great angles on the Republican nomination. This is really rare. Most candidates do not even have one. Fred actually had two.

First, he hit every item on the conservative checklist - without contortions. In this year's Republican field - with Giuliani, Huckabee, McCain, and Romney - this was a boon. Fred was no apostate. He was the "consistent conservative."

Second, he pulled off a great coup in the late summer. This year's frontrunning candidates thought that, to get enough cash to compete, they had to begin campaigning in January of last year. They had to hold rallies, release endless position statements, and help the media perpetuate the fiction that they were actually persuading "undecided" voters. Donors, you see, have come to expect candidates to entertain them with the trappings of a campaign, even if there was no real core to it all.

But Fred found a way around this inanity. He correctly figured that if he was the last candidate to enter the campaign - he would garner so much instant attention that he could compete as if he had been campaigning for a year.

This is the sort of "cheating" that makes a rat choice guy weak in the knees. To appreciate just how clever this was, consider this analogy. These days, telephone companies offer unlimited packages - you pay a flat fee and you call as much as you want. Suppose that you know that all of your friends and family had these packages - what is the rational thing to do? Get the cheapest plan you can, and let your friends call you! Fred did something like this. The endless campaigning of all his competitors created an environment that would be conducive for a late entrant to steal the limelight. His competitors created it with their time and money. Fred, by coming in late, could reap all of the rewards without paying the cost.

I was mightily impressed by this move. I still am.

All in all, this was a great place to be. Fred was so well positioned when he entered the race that Romney took a shot at him in his first debate: "[The series of debates is] a lot like 'Law and Order,' Senator. It has a huge cast, the series seems to go on forever, and Fred Thompson shows up at the end." Congrats, Fred. When Romney hits you - you've made the big time.

So what happened? Fred squandered this incredible opportunity. Lackluster speeches, lackluster debate performances, lackluster campaign stops. We saw it happen in slow motion through the fall - the drip-drip-drip of stories about him half-assing his campaign appearances corresponded with his plummeting poll positions. Apparently, Fred just doesn't like to campaign - and the voters took notice. Forget the hyper-kinetic campaign of 2008, where candidates make endless pleas for votes. Fred probably would have disliked McKinley's front porch campaign of 1896. "What...another group of people are out front? Give me a break!"

Personally, I don't hold this against him. Campaigning seems like it would be a drudgery, doesn't it? Watching C-SPAN on the weekends has been instructive for me. They show unedited feeds of campaign speeches, and the glad-handing that follows. I would not want to do what those candidates do. They give the same speech day in and day out. They facilitate the traveling press corps - a pack of jackals who would delight in tearing them to shreds. They talk endlessly with voters, which I wager is more of a chore than it might first appear. Americans are mind-bogglingly opinionated when it comes to politics. We are convinced of the veracity and rectitude of our preferences, and we have no problem giving others an earful of our consequential utterings. Especially candidates. It's our right, damnit - and we have no bones about exercising it.

Fred obviously had no taste for any this. So, he half-assed it. He didn't campaign in a new way, via new media and all that. Instead, he campaigned the old way, but did a lousy job by it. Eventually, we all caught on. And his numbers dropped. And dropped. And dropped.

And one day late last year, he realized that voters no longer saw him as viable. They agreed with him on the issues, for sure - but they weren't going to vote for him. He went from frontrunner to another victim of the Catch-22! He responded by campaigning non-stop - but by then, he was already in that horrible circle. He couldn't win over voters because they thought he couldn't win over voters. That is how a man of his conservative bona fides could meet his end in South Carolina.

Or as an astute emailer put it: "I was a Thompson supporter. Unfortunately, I supported him more than he did."

All in all, I think Fred embarrassed himself a little bit. Not by running and losing. Lots of people run for the presidency, lose the race, and retain their reputation. But his decision to run was ill-advised. He did not have it in him - and he wrongly thought he did.

Final point. While many things are possible in politics - I think it is extremely unlikely that Fred will be chosen as a vice-presidential nominee. Can the Republican nominee be certain that Fred will do what it takes to win? Of course not. Sure, Fred could help bring the party together. But lots of people can do that from a vice-presidential spot. So, why take the risk?

-Jay Cost

Is McCain Inevitable?

James Madison had nothing to do with the Republican nomination system. Nevertheless, it bears his imprint. Madisonianism is deeply ingrained in the American political consciousness. When we invent new political institutions, as both parties did in the 1970s when they redesigned their nominating schemes, our inventions often bear the mark of the Father of the Constitution.

Madison saw the government as a mediator of factions, none of which can be expected to have the national interest at heart. Therefore, a true republican government must prevent one faction from imposing its narrow interests upon the whole. This is why our system has so many "checks and balances." They enable one faction to stop another when its interests are threatened. Madison's thinking was that this kind of system would ensure that the national interest was never sacrificed for factional gains.

The rules governing how the Republican Party selects a presidential nominee reflect this Madisonian objective. Consider:

(1) A broad base of people may factor into the nominating system. In some places, anybody can vote. In other places, only Republicans can.

(2) Voters have an important, but limited, role in selecting the next nominee. They do not directly choose the nominee. Instead, they choose delegates who choose the nominee (in some states, they choose delegates who choose delegates who choose the nominee!). In most cases, delegates only have limited obligations to the votes of their states. Many delegates have no obligation at all.

(3) The manner of choosing delegates is diverse. Some are chosen based upon the primary/caucus results in congressional districts, and some based upon statewide results. Some are selected by winner-take-all rules, some by proportional rules, some by state conventions, and some by virtue of their role in the RNC.

(4) A candidate must win an outright majority of delegates to become the nominee. Nobody can win by appealing to a narrow slice of the party - be it ideological, demographic, or geographical.

(5) This is largely a bottom-up system. State parties and governments have the biggest say in how delegates are chosen - the RNC simply establishes basic guidelines.

So when we look closely, the seemingly quirky nature of the nomination process reveals its Madisonian roots. Candidates need to build broad coalitions to win. When they fail to do that, disenchanted minorities can slow or even stop the nomination of a candidate who lacks universal appeal.

Now that he is the frontrunner, this is the problem that confronts John McCain. In every previous cycle in the modern era - the Republican who wins South Carolina wins the nomination. A big reason is that the victory in the South, the heart of the Republican's general election strength, signals who the favored candidate is. The rest of the candidates eventually recognize this, and they bow out. McCain won South Carolina, and he is better positioned now than he was a week ago - but the race is not over.

McCain is staunchly opposed by a vocal group of conservatives who view him as an unreliable maverick. You can hear their most prominent advocate on the radio every weekday from noon to three eastern. You can see them in the exit polls, which show that McCain has not yet won a (statistically significant) plurality of Republican voters, nor those who consider themselves "very conservative." In years past, opposition to the Republican frontrunner tends to fade away after South Carolina, with the supporters of the loser accepting that their guy can't prevail and reconciling themselves with the victor. But that does not seem to be happening this year. There is a faction of the party that seems unwilling to accept McCain. It might be able to stop him.

It should be clear from the nomination rules that somebody could find enough delegates to oppose McCain on the convention floor - even if he did not offer a serious challenge early in the process. From the unpledged delegates, to the delegates allocated by conventions, proportional allocation, and the congressional district delegates - there are a lot of ways to win convention support even as somebody else "wins" states. Eventually, an opposition candidate would have to break through with outright victories. He cannot win the Republican nomination underground - but the way delegates are allocated could keep the race close until he breaks through. Importantly, about 65% of South Carolina voters preferred somebody other than John McCain. This tracks with his standing in the national polls. So, the anti-McCain faction might have an audience - if it can find a candidate to rally behind. Also of importance: 95% of all delegates have yet to be allocated. And even after Super Tuesday, 45% will remain to be allocated. The faction has time to make its case.

I am not saying it will be successful. McCain has a very strong chance to win the nomination. One feather in his cap is that opposition to him does not cut cleanly along any ideological line. Rick Santorum is vehemently opposed to him, but Tom Coburn just endorsed him. Another asset is that the Republican delegate allocation system is much less charitable to losers than the Democratic scheme - this gives the opposition less time to get its act together.

My point is simply that the opposition to McCain could prove to be important. For better or worse, the old maverick from Arizona has inspired intense opposition in some quarters. In a nomination system such as this - that opposition might ultimately be able to stop him.

With McCain as the frontrunner - the way to look at this nomination battle should shift. Most of us had written McCain off last summer - so we were not expecting him to precipitate an ideological battle. If anything, we were expecting some kind of bottom-up opposition to Giuliani - with party elites accepting his candidacy, and rank-and-file pro-lifers rejecting it. The rise of McCain scrambles all of this. There is an ideological conflict brewing in the GOP - but not the one we thought we would see. This means that the way we have looked at nominations over the last few cycles does not hold. I think this contest could be longer than many have intuited - and the results in Florida could determine exactly who emerges as the "anti-McCain" candidate.

Do not expect the press to catch this dynamic. It understands the here-and-now of contemporary politics much better than the forces and institutions that have guided it for decades. One effect of its misunderstanding will come on Super Tuesday, which it will treat just like the general election. That evening, it is going to focus relentlessly and exclusively on who wins which states - as if delegates are allocated like Electoral College electors. Do not get caught up with this, regardless of how splashily it is staged. With the prospect of a McCain candidacy, and the ideological divergence it implies - this is not the best way to analyze Super Tuesday, even though it is an important aspect. We also need to wait until the next day to see how the delegates are meted out - that will indicate just where this race is going to go.

If McCain were a consensus candidate - like Bob Dole in 1996 or George W. Bush in 2000 - I would say that his victory in South Carolina, combined with his lead in the national polls, would be sufficient for the nomination. Florida would offer a final validation - and that would be it. But McCain is not this candidate. He has serious, entrenched opposition - and in a system such as ours, entrenched oppositions are given opportunities to stop something from happening. I do not know if it can stop McCain, but I expect it to try.

-Jay Cost

On Yesterday's Results

Democrats in Nevada

How did Hillary Clinton win Nevada? We'll look at the Nevada entrance polls to get a sense. Note that these show a slightly larger margin of victory than the results you have seen on television and the Internet. There are three reasons. First, the published results are really counts of delegates to the state convention. The Nevada Democratic Party is not releasing actual vote counts. Second, there was a 15% viability threshold that might have influenced delegate counts; for instance, caucus goers who came to vote for John Edwards might not have had a chance to win their guy some delegates. Third, the entrance poll is exactly that - a poll. It has a margin of error just like any other.

With these caveats, there are some interesting observations we can make about the Nevada results. Generally, Clinton won in Nevada because she retained the voting coalition that she formed in New Hampshire.

Consider:

- Once again, Clinton carried a strong majority of female voters, who also constituted an overwhelming majority of caucus-goers. She won them 51% to 38%, and they made up 59% of the vote. This basically mimics her success in New Hampshire.

- She expanded her lead among Catholic voters. In New Hampshire, she won them by 17 points. In Nevada, she won them by 27 points. In Nevada as in New Hampshire, Obama and Clinton split Protestants.

- She won voters who consider the economy their top concern by 9 points in both Nevada and New Hampshire. She once again won a solid victory among voters who make less than $50,000. In New Hampshire, it was by 15 points. In Nevada, it was by 12 points.

As she won the vote by a larger margin in Nevada than she did in New Hampshire - it stands to reason that she expanded upon this basic voting coalition. Indeed, she did. Here is how:

- She cut into Obama's share of the male vote. Obama defeated Clinton among men by 11 points in New Hampshire. In Nevada, he beat her by just 2 points.

- She did well among Hispanic voters, winning them by 38 points. But Obama won African Americans by 69 points. Interestingly, Hispanics and African Americans both constituted 15% of entrance poll respondents. So, Clinton's win among Hispanics was more than matched by Obama's win among African Americans.

- Older voters seemed to have comprised a larger share of the vote in Nevada, and Clinton won them by a larger margin than she did in the Granite State. Voters 60 and older were 36% of the electorate in Nevada, and Clinton won them by 29 points. In New Hampshire, voters 65 and older made up 13% of the vote, and Clinton won them by 16 points.

- Clinton won a much larger share of white voters. In New Hampshire, she won them by just 3 points. In Nevada, she won them by 18 points.

- Clinton improved her totals among voters who make more than $50,000 per year. Obama won them by 5 points in New Hampshire; Clinton won them by 5 points in Nevada.

Once again, it appears that Hillary Clinton won by turning out a traditional Democratic voting coalition: Catholics, women, and "downscale" Democrats. This time, she added to this coalition with strong showings among Hispanics, whites, men, and "upscale" voters.

Nevertheless, there is evidence that Obama is able to take a solid portion of the core Democratic vote - notably African Americans. This is good news for Obama in the short term. If you take these demographic preferences to South Carolina, Obama will probably win because each group's share of the vote shifts. For instance, Hispanics are not a major factor in South Carolina, and African Americans are a much greater factor. This alone would probably yield Obama a victory next week.

But in the long run, my feeling is that a replication of Nevada's result would give Clinton great success on Super Tuesday. The real concern for Obama should be the shift of white voters to Clinton. It remains to be seen whether this is sustainable (we saw nothing like this in Iowa or New Hampshire). If it is, Obama is in real trouble.

Final point. I mentioned in my write-up of the New Hampshire results that we do not yet have a handle on union voters. They broke for Clinton in New Hampshire, but they split equally between Clinton and Obama in Iowa. Nevada reflects the Iowa results. Clinton won voters from union households by 9 points in New Hampshire. She won them by just a point in Nevada. Much of the difference here surely has to do with Obama's endorsements from big Nevada unions - but that is precisely the point: we don't know whether, or even if, union voters are going to break one way or the other on Super Tuesday.

Republicans in South Carolina

If demography has been key in the Democratic races, ideology seems to be the difference maker in the Republican contests. This should come as no surprise, given that the differences among Democrats have been tonal and the differences among Republicans have been issue-based.

How did McCain win? Simple. He did exactly what he did in New Hampshire. Namely:

- He won a big victory among voters who disapprove of the Bush administration. In South Carolina, he won them by 13 points over Huckabee. In New Hampshire, he won them by 14 points over Romney.

- He put together a coalition of moderate and slightly conservative voters. He won moderates by 30 points in South Carolina, and by 17 points in New Hampshire. He won those who call themselves "somewhat conservative" by 2 points in South Carolina, and by 3 points in New Hampshire. Huckabee won those who are "very conservative" by 22 points in South Carolina; Romney won them by 25 points in New Hampshire.

- He split self-identified Republicans with Huckabee. He won 31%, Huckabee won 32%. In New Hampshire, the results were basically identical. Romney won 35%, McCain won 34%.

- He won self-identified Independents by 17 points in South Carolina. In New Hampshire, he won them by 13 points.

Huckabee's strength was clearly among self-identified evangelicals. They made up 55% of the electorate, and he won them by 16 points. Unfortunately for him, he lost non-evangelicals by 25 points. He actually finished fourth among non-evangelicals, behind Romney and Thompson (though the differences among the three are statistically insignificant). This is a sign that Huckabee's campaign has not successfully expanded beyond the core voters that it first wooed in Iowa. The following was a telling indicator of that. Despite the fact that Huckabee has run on a "populist" message, he and McCain split voters who identify the economy as their biggest concern, 32% to 32%.

Another point. It is hard to argue that a more vigorous campaign by Romney would have stopped McCain. It appears that Huckabee took over the position Romney had in New Hampshire with strong conservatives and Bush administration approvers. As a matter of fact, Huckabee's loss was probably due in part to the fact that Thompson was in the race. Thompson did relatively well among those who support Bush and among the "strong conservatives." Adding another candidate who makes the same appeal probably would have only helped McCain.

What we thus have in the Republican Party is the makings of a plurality coalition in which a prominent portion - namely, strong conservatives - is not fully a part. McCain has managed to win twice even though he has not won over strong conservatives. Can this continue? I am not sure - though I surely think South Carolina's confirmation of the New Hampshire vote strengthens his hand.

It will be interesting to see how high-profile conservative leaders react. I am sure they will not take this victory as the end of the race. They can hang their hats on this: if there had been a single "conservative" candidate in South Carolina, he might have defeated McCain. There was a split between McCain, Huckabee, Romney, and Thompson among "very conservative" voters. If there had been just a single "conservative" candidate to oppose him - McCain might have lost.

-Jay Cost

Can Mitt Catch On?

Romney's win in Michigan keeps his candidacy alive - but what are the odds that he will win the nomination? I'm not sure, but they are much reduced from what they were six months ago.

Early last year, the Romney campaign put together a plan that - if it panned out - would probably have won him the nomination. The idea was for Romney to build a huge war chest that would enable him to compete everywhere. He would then win Iowa and New Hampshire, emerge as the consensus Republican candidate, and overwhelm the rest of the field.

But the plan backfired. Romney lost Iowa, and then he lost New Hampshire. Accordingly, he is not the consensus candidate of the party. Far from it. While he has a toehold in the GOP electorate, that's all he has. The recent Pew poll offers cross-tabs that tell the story in vivid detail. Even though the poll was completed before the Michigan primary, there is still a good bit to learn from it:

Pew Poll.jpg

Huckabee's strength is with evangelicals. McCain's strength is with self-identified moderates and liberals; he is also strong among mainline Protestants and Catholics. Romney wins a solid portion of self-identified conservatives - but he is in a three-way statistical tie with Huckabee and McCain for their support. Clearly, he has not yet broken through with either demographic - be it ideological or religious. You could also slice the party by income - and you would see the same result. McCain dominates; Huckabee has a fair share; Romney has not broken through.

I think this is why Romney is skipping South Carolina. With the current alignment of the vote - Romney cannot expect to win a sufficient slice of the South Carolina electorate. His victory in Michigan might have shifted things in his favor (we'll know in a few days) - but obviously not enough to put South Carolina in play for him.

So, what is Romney's angle on the nomination? He heads to Nevada and wins that state's uncontested caucus. This keeps him viable until Florida, regardless of what happens in South Carolina. He then gives Florida everything he's got.

Will it work? I don't know. He has another potential problem.

Why is it that most primary candidates refuse to run sustained, intense negative campaigns? The answer is that everybody is basically on the same side. An attacking candidate has to be careful about his opponent's core supporters. He runs the risk of alienating them - and they might ultimately refuse to support him after their guy drops out of the race. Romney might find himself in that situation. His attacks on McCain and Huckabee have been as sustained and intense as any this cycle. And there is evidence that this has damaged him with the Mac and Huck factions.

The Pew poll found that Romney's net favorable rating among these voters is not very strong: just +7% among McCain voters, and a whopping -9% among Huckabee voters. Of course, the sample sizes informing these statistics are small - but they are large enough to validate this modest conclusion: Romney is relatively weak among Huckabee and McCain supporters. For comparative purposes: McCain is +30% among Huckabee supporters; Huckabee is +15% among McCain supporters; Giuliani is an eye-popping +69% among McCain supporters, and +33% among Huckabee supporters. [A problem Romney will confront if he wins the GOP nomination: he has a net -12% favorable rating among the general electorate. I'd wager this is also a consequence of the negative tenor of his campaign in recent months.]

This could create problems for Romney in Florida, depending on how things turn out in South Carolina. Following Pew, it does not seem that Romney is the second choice of a plurality of Huckabee voters or McCain voters. The situation in Florida might be different than what Pew finds on the national level, but I doubt it is significantly so. My sense is that if Floridians bolt Huckabee after he loses South Carolina - a plurality will go to McCain, not Romney. Similarly, if they bolt McCain - a plurality will go to Giuliani, not Romney. Generally, Pew and other pollsters have found Romney in third or fourth place when it comes to second choices. Pew also finds that 20% of Republicans will never vote for Romney, making him more "unacceptable" than McCain or Giuliani.

In light of this, I think that what Romney needs is a nominal Huckabee (or Thompson) victory in South Carolina. It would keep the field as open as possible. If the Florida electorate is split four or five ways, Romney might be able to pull out a victory based on his current coalition - thus giving him an opportunity to expand it in advance of Super Tuesday.

Ironically, Giuliani is in the same position. This is one of the strangest features of this year's race. Romney has competed everywhere, Giuliani has competed nowhere - but they are both stuck in the middle and in need of an open field.

-Jay Cost

On the Michigan Results

Romney's victory last night was decisive. Let's dig into the exit poll results a little bit to see how he pulled it off.

First off, Romney won in Michigan the same groups he won in New Hampshire. The difference was that he won them by wider margins. Consider:

- In New Hampshire, Romney won voters who support the Bush administration, 37% to 32% for McCain. In Michigan, he improved upon that margin, 45% to 24%.

- In New Hampshire, he split Republicans with McCain, 35% to 34%. In Michigan, he won them decisively, 41% to 27%.

- Independents went for McCain last night, 35% to 29%. But they went for him more strongly in New Hampshire, 40% to 27%. Plus, Independents made up just 25% of the vote last night, compared to 37% in New Hampshire.

- In New Hampshire, Romney won those who identified themselves as "very conservative," 43% to 18%. In Michigan, he won them, 48% to 11%.

- Romney improved upon his support among those who identified themselves as "somewhat conservative." McCain won them in New Hampshire, 38% to 35%. In Michigan, Romney won them, 35% to 32%.

But not all the news is fantastic for Romney.

Last night, the McCain campaign was spinning its loss as a consequence of the fact that Romney is a "favorite son." There is evidence to support this claim. 42% of respondents in the exit poll reported that Romney's ties to the state were important factors in their vote choice. Romney won those voters decisively, 58% to 17%. Meanwhile, 56% of voters said that his ties to the state were not important factors. Interestingly, McCain won those voters, 39% to 23%.

I think this data should induce restraint. We should be careful not to over-interpret these results. Clearly, Romney's win was decisive enough to keep him in the race. But it is also clear that he received a sizeable boost from his ties to the Wolverine State. How much of a boost - we cannot know for sure. Accordingly, we cannot say whether these results are an omen of things to come.

Another factor of critical importance in Romney's victory was the economy. Romney's message in the last few days was tailored to the economic concerns of Michigan. It seemed to work. In New Hampshire, McCain won voters who said the economy was their most important concern, 41% to 21%. In Michigan, Romney won them, 42% to 29%.

This probably gives the Romney team some insight on what to say next - perhaps this economic message will work in the other states. John McIntyre's essay from December seems prescient to me now: with Huckabee on his right flank, and McCain on his left - Romney can run as the down-the-middle conservative. His message? The economy.

Of course, Michigan voters were more concerned about the economy than others. For instance, 31% of New Hampshire exit poll respondents listed the economy as their primary concern, but in Michigan it was a whopping 55%. And, what is more, Romney's appeal on the economy might be limited to Michigan. He ran on an explicit promise to revitalize the auto industry.

I wonder if there is a connection here. Does Romney's specific, auto industry pitch connect his Michigan ties to his strength on the economy? Was the average Michigander thinking, "I'm concerned about the economy - and I'm voting for Romney because I know he won't forget us!" This would be an easy hypothesis to test if the networks provided raw data - which they don't. So all we can do is speculate.

We can approach the economy issue from another angle. According to the exit poll, Romney won areas outside metro Detroit, 35% to 33%. He won metro Detroit, 42% to 29%. Most of his strength came in Oakland and Macomb counties. The exit poll reports that he took them 46% to 28%. In 2000, McCain beat Bush in Oakland and Macomb - 50% to 45%. Metro Detroit is the area hardest hit hit by the auto slump - and thus most susceptible to Romney's economic appeal.

What can we conclude from this? Clearly, Romney won a decisive victory. Just as clearly, some portion of this win was due to his special connection with the state as well as its special economic circumstances. Unfortunately, it is not possible to quantify exactly what portion this was. Thus, I think the fair conclusion is that Romney's win keeps him viable moving forward, but it probably does not give him any special advantage. As Tom Bevan has pointed out on the RCP Blog - the same can be said for Huckabee's win in Iowa and McCain's win in New Hampshire. They were, in many respects, consequences of "specialty" appeals that are not necessarily generalizable. I think the same goes for Romney's victory last night.

A final point on the Republican results. The exit polling data offers some counter-intuitive evidence about who is being helped by the conduct of the Iraq war. In New Hampshire, Romney won voters who approve of the Iraq war, 37% to 33%. McCain won those who disapprove, 44% to 19%. These basic results were replicated in Michigan. Romney won approvers, 42% to 27%. McCain won disapprovers, 36% to 29%.

These results are, I think, a blow to the idea that McCain did so well in New Hampshire because of the surge in force levels. They also may foretell trouble for McCain in South Carolina - even though he was an early and strong supporter of the surge, he does not seem to be getting a great deal of credit for it.

As for the Democratic side - the big story is Hillary Clinton losing the African American vote to "uncommitted." The exit poll pegged African Americans going against Clinton, 68% to 30%. It appears that opposition by African Americans induced a split in Wayne County (where Detroit is), 50% to Hillary, 45% to uncommitted. People in the media are going to connect these results to the racial kerfuffle of the last few days - and they are partially right to do so. But I think there is more to it than this. Since his Iowa victory, Obama's numbers among African American voters have been trending upward. Tonight's results are another indication that African Americans are breaking his way. The Clinton campaign should be worried about this. It appears as if Obama might be able to take an important part of the traditional Democratic coalition. He is thus moving beyond the relatively narrow appeal of previous "insurgent" Democratic candidates like Bill Bradley and Gary Hart. This is bad news for Clinton.

-Jay Cost

Looking ahead to the Republican Convention

As the Republican race is still wide open, I thought it time to review how delegates to the Republican National Convention will be allocated this year. There has been talk about a "brokered convention." And while I think that there will probably be a clear nominee come September 1, 2008 - the probability of a convention fight is much greater than it has been in the last thirty years. Accordingly, we are well advised to get a lay of the land, to see exactly how the GOP convention might proceed if there is no clear winner.

We will look at three relevant features: how delegates to the convention are chosen, what are the factors that could cause a brokered convention, and how a deal could be worked out if there is no clear nominee.

Delegate Selection

First off, the Republican presidential nominee is chosen by delegates who vote by ballot. These delegates have been selected to attend the national convention - most of them are directly determined by state primary and caucus results. To win the nomination, a candidate must win a majority of the 2,380 delegates in attendance. That means the first to get 1,191 delegates wins.

There are four types of delegates to the Republican National Convention. The first type consists of elected members of the Republican National Committee (RNC). Each state is allocated three such delegates. Second, each state gets three delegates for every congressional district. Third, each state gets ten "at large" delegates. Fourth, each state gets "bonus" delegates added to their at large delegation based upon how Republican the state has voted recently. [Note that United States territories like Puerto Rico and Guam also receive at large and RNC delegates.]

So, put all of these rules together - and we get a result like the following.

California has 173 delegates to send to the convention. That is:

- 3 delegates from the RNC

- 10 at large delegates

- 53 congressional districts X 3 delegates per congressional districts = 159 district delegates

- 1 bonus delegate because it reelected Governor Schwarzenegger in 2006

Pennsylvania has 74 delegates to send to the convention. That is:

- 3 delegates from the RNC

- 10 at large delegates

- 19 congressional districts X 3 delegates per congressional district = 57 district delegates

- 1 bonus delegate for U.S. senator + 1 bonus delegate because a majority of its House delegation from 2004 through 2007 was Republican + 1 bonus delegate for state house control + 1 bonus delegate for state senate control = 4 bonus delegates

A convention would be "brokered" if, on the first ballot, no candidate wins a majority of delegates. In that situation, another ballot would be taken to find a majority winner. And another ballot and another and another until one candidate has won a majority. The term "brokered" refers to the need for a settlement to be brokered among candidates, delegates, and party leaders. If no candidate is the majority winner, a deal will need to be worked out to induce some delegates to change their votes.

What Could Cause a Brokered Convention?

Obviously, the critical factor for a brokered convention is a situation in which no candidate has a majority of delegates. This is why I would wager the Republicans are more likely to have a brokered convention than the Democrats. If delegates are split in just two ways, between Clinton and Obama, then there is only one situation in which a brokered convention could occur: a strict numerical tie. Of course, if Edwards has delegates - there would be other ways for a stalemate to occur; however, the fewer delegates he has won, the less likely such a situation would be. If, on the other hand, delegates are split between Rudy Giuliani, Mike Huckabee, John McCain, Mitt Romney, and Fred Thompson - a brokered convention is much more likely to occur.

Another important factor is how delegates are allocated. An interesting development in the Republican Party over the last few years has been a movement from winner-take-all delegate allocation schemes to proportional allocation. This makes a big difference - in a winner-take-all contest, the winner of the primary or caucus wins all of the delegates. The losers win nothing. Accordingly, the probability of a brokered condition decreases as the gap between winners and losers increases.

This year, however, the Republican Party has twelve states where at large and congressional district delegates are chosen by proportional representation. What is more, seven states have a mix of winner-take-all and proportional rules. For instance, Alabama's delegates are allocated winner-take-all if the winner receives over 50% of the vote. If he does not, the delegates are allocated proportionally. Meanwhile, twenty-one states plus the District of Columbia allocate their at large and congressional district delegates on a winner-take all basis, and eleven states select their delegates at state conventions or something similar. This means that, in many states, candidates who lose the statewide vote can win delegates - thus giving them a better chance to scrap together enough supporters to make a run at the nomination.

A third factor that could contribute to a brokered convention is the fact that many states allocate congressional district delegates based on who won the congressional district - not who won the state. Ohio, for instance, allocates its 10 at large and 21 bonus delegates based on who wins the most votes in Ohio. But it allocates its 54 congressional district delegates based on who won which congressional districts. So, once again, a candidate who lost the statewide vote could still win a few delegates from Ohio because he won congressional districts.

Resolving the Dispute

Suppose that these factors make it so that no candidate wins a majority of delegates. What would happen next? Obviously, a deal would have to be worked out between candidates, party leaders, and delegates - but the contours of such an agreement would probably not resemble the ones brokered in the "smoke-filled" rooms. In the modern era - many delegates are bound to primary and caucus votes. That is, they are sent to the convention forced, in one way or another, to vote the way their state voted.

Out of the 2,380 delegates sent to Minneapolis St. Paul - 1,729 of them will be bound in some formal way (this figure excludes Ohio, Washington, North Carolina, and the Virgin Islands, whose delegates are "morally bound," "unofficially bound," or "requested" to vote for their candidate). These break down in the following way:

- 463 delegates will be bound through the convention.

- 565 delegates will be bound through one ballot. That is, they have to follow the results of the state election on the first ballot. After that, if no candidate has a majority of delegates, they are free to vote as they please.

- 383 will be bound through two ballots.

- 318 will be bound through three ballots.

The remaining 651 are not bound in a formal way. They can vote however they want from the first ballot.

What effect will this system have? My intuition is that it might make it harder to resolve a dispute prior to the convention. The Republicans will finish allocating delegates sometime in late June. So, we should know by then if no candidate has won a majority of delegates. However, how could the various factions work out an agreement that can be implemented before the convention? They might be able to - but it will be harder because many delegates actually have to cast at least one ballot for their initial candidates (unless they are released by a candidate who withdraws - and it is unlikely that a candidate with a significant batch of delegates would withdraw before the convention). What is more, candidates might be interested in waiting to see how the unbound delegates behave in the first few ballots to get a sense of their full strength. They might also be willing to wait until the fourth ballot to see if they can pick up any of the formerly bound delegates. All of this could complicate a pre-convention deal in a whole host of ways: strategic candidates might prefer to wait for a few rounds of balloting to see where they stand.

Another complicating factor is that patronage is no longer an emollient. For instance, how would you "buy off" 500 or so die-hard Romney delegates? In this age, the party lacks patronage resources to give them consolation prizes. And, for that matter, Romney delegates would probably not want anything except a Romney nomination. It is unclear to me how one candidate could negotiate any deals except by offering the vice-presidency to one of the losers - which probably would do little good if the delegates are split roughly equal among three or more candidates. It is easy to envision a cycle. The Romney people buy off the Guiliani people by making Guiliani the vice-presidential nominee. That leaves the McCain people out in the cold, so they offer Guliani the presidential nomination if he makes McCain the secretary of state. The Romney people respond by offering McCain the top spot if Romney can have the veep position. And so on.

All in all, if there is no clear delegate winner in the summer, and a deal needs to be brokered - I expect it to look very different from those hammered out in the classic brokered conventions of the past.

-Jay Cost

How Do You Solve a Problem Like McCain?

Update, 1/21/08: John McCain's win in South Carolina has placed him in a strong position to win the Republican Party's nomination. However, he still faces an important obstacle: opposition from conservative leaders. I think this opposition could create an interesting dynamic over the coming weeks. Accordingly, I thought it worthwhile to republish this essay, which was originally offered last week.

***

Since his win in New Hampshire, many have come to view John McCain as the Republican frontrunner. He has a lead in the RCP national average, and he is the favorite at InTrade's future's market.

I think this talk is a bit hasty. Most obviously, McCain and Romney are tied in Michigan - and the polls over the last three weeks have been largely conditioned by who has won. So, who knows how the Michigan results will influence South Carolina and the rest of the nation.

There is another reason I am hesitant about this bandwagon. I think that the problem that nearly destroyed his candidacy last summer is still there - and it could yet do him in. The problem? Conservative leaders do not care for his candidacy.

Before I get into this, I need to clarify a common misconception. Many people wrongly assume that average voters are as ideological as party leaders (broadly defined to include elected party members, prominent personalities in the party, and ideologues on op/ed pages and in opinion journals). They are not. In fact, the highest estimate I have seen for the percentage of the public that thinks ideologically is 30%. And many of those people think in an effectively ideological manner - e.g. they use the information they get from trusted opinion makers to formulate opinions on inter-related issues.

This means that McCain is probably more acceptable to rank-and-file Republicans than we might initially think. For instance, according to the New Hampshire exit poll, McCain won voters who identified themselves as "somewhat conservative," 38% to 35%. Romney won those who identified themselves as "very conservative," 43% to 18%. These results are consistent with the previous paragraph. Romney won the strong ideologues. McCain won the weak ideologues.

McCain's problem has more to do with conservative leaders than the rank and file. You probably have picked up on this impressionistically. National Review endorsed Romney. Rush Limbaugh has been openly critical of McCain's candidacy. Human Events is no longer the significant intellectual force it was three decades ago - but I thought it was telling that it endorsed Thompson after McCain won New Hampshire.

If you examined McCain only on paper, this would surprise you. By many metrics, he is the classic Republican frontrunner. He has been in Congress for more than 25 years. He was on Bob Dole's vice-presidential short list. He was the runner-up in the 2000 contest. He has a lifetime American Conservative Union (ACU) rating of 82.3. Above all, he has great credibility on two issues that will be important in November, the Iraq war and the war on terror.

Nevertheless, opinion makers do not much care for his candidacy. Apparently, neither does the Republican congressional caucus. Examining congressional endorsements for McCain and Romney, excluding in-state supporters (as they often act more out of home state pride than ideological proximity), we find some interesting results. 34 Republicans have endorsed Mitt Romney, while just 24 have endorsed McCain. Furthermore, Romney's supporters are more in line with conservative opinion. Their average 2006 ACU rating was 84.1, and 26 of them come from states Bush won in 2004. Meanwhile, the average 2006 ACU rating for McCain's supporters is 70.7, and just 12 of them come from Bush states. In light of McCain's résumé, this is consequential. He should have locked up most members of the Republican caucus, but he has not.

I believe that McCain's long-standing disagreements with conservative leaders could hurt him moving forward, just as it hurt him in the summer. One of his big problems is that viability is partially a subjective assessment. It has a complex relationship with likeability: people tend to view candidates they like as more likely to win the nomination, and they tend to like candidates the more they think they can win the nomination.

I think this contributed to the McCain "implosion" last July. The demise of his candidacy was quite overstated. Lots of campaigns have staff shakeups and money problems. But McCain's troubles were taken as a sure sign he was finished. I think that shows how likeability can affect viability. Conservative leaders were more inclined to write McCain off because they do not like him. They stopped talking about him, and the voters stopped considering him a viable alternative.

As we all know, McCain stuck it out - and New Hampshire delivered him a big win. He is now back at the top of the polls. But, I wonder whether he will once again be undone by the tense relationship with conservative leaders. I see two potential problems.

First, should McCain lose Michigan on Tuesday, he might be criticized more harshly than a mainstream conservative candidate would. Look, for instance, at Romney - who, despite having lost Iowa and New Hampshire, is still considered viable by many leaders. It is unlikely that McCain would enjoy that kind of charity should he lose Michigan or any subsequent state. This gives him a smaller margin for error.

Second, I think the prospect of a McCain nomination gives his mainstream opponents a wider margin for error. I do not think that McCain can end the nomination battle with wins in Michigan and South Carolina. Because conservative leaders dislike him, I think they might "rediscover" a viable alternative to him. Romney, Thompson, or even Giuliani could be brought back to life in a final effort to stop McCain. This would not happen if he was more acceptable to conservative leaders. Again, I think this is a big part of what is keeping Romney alive right now. No candidate in the modern era has made such a try for Iowa and New Hampshire, lost them both by wide margins, and gone on to win the nomination. The fact that many leaders are still holding out hope that Romney can win is, I think, a testament to how much McCain rubs them the wrong way.

By themselves, conservative leaders cannot stop McCain. The liberalization of the nomination process has left the upper echelons of both parties less able to determine directly who wins. However, they have indirect power. They can influence which candidates are seen by the voters to be credible candidates. Through their dialogue with one another as well as their direct communications to the public - they help establish voter expectations, and therefore the range of viable alternatives voters perceive. The more they talk up a candidate's viability, the more viable he becomes. The less they talk it up, the less viable he becomes. This is the power to set the agenda. Conservative leaders could help take McCain off the agenda if he falters, or they could help to place an opponent on the agenda to stop him.

Final point. Everything that has been written here applies to Mike Huckabee all the more. Huckabee is even more unacceptable to conservative leaders (he has the endorsement of only three out-of-state Republican members of Congress). His campaign even seems to be courting their contempt. McCain, to his credit, has worked hard since the immigration bill not to thumb his nose at conservative leaders. But not Huckabee. And I think that they would only accept Huckabee's nomination after he has won a majority of delegates. Until then, they'll find somebody, anybody to oppose him.

-Jay Cost

Clinton v. Obama: Moving Forward

I have been doing some digging into the numbers out of Iowa and New Hampshire - and the data I have found indicates that the Democratic nomination battle could be extremely tight.

It is common for scholars, academics, and number-crunchers to review vote results by demographic category - and to construct a story about how the winner actually won. I engaged in a bit of this on Wednesday, arguing that Clinton won New Hampshire by pulling together a traditional Democratic voting coalition. This is how most frontrunning Democrats have won the nomination - and, so I labeled it the "FDR Coalition" or the "Mondale Coalition." This is the coalition you learned about in your civics classes. Union workers, Catholics, African-Americans, lower-income voters, etc. Clinton pulled a good-sized chunk of these voters together, combined it with her overwhelming success among female voters, and got a solid win in New Hampshire.

A very impressive one, I hasten to add. New Hampshire is a quirky state. It has a habit of voting for insurgent candidates who appeal to more upscale, educated, professional class Democrats. I'm thinking in particular here of Gary Hart, who won New Hampshire in 1984. For Clinton to win in New Hampshire with the "Mondale Coalition" is something that Mondale himself did not do!

A quick review of the upcoming states can give us a sense of just how powerful this coalition can be.

The following table reviews the union members by percent of working population in the Super Tuesday states, plus Iowa and New Hampshire thrown in for comparative purposes. It excludes New York and Illinois, presuming that they will go heavily for their home state candidates. It also lists "pledged" delegates, i.e. those delegates who are allocated based upon performances in the primaries/caucuses. Finally, it includes an asterisk to identify caucus/conventions, and a plus sign to identify closed primaries not open to Independents.

Union.jpg

As you can see, unions will matter on February 5th. Remember that union members tend to be Democrats. They also tend to be reliable voters. So, the actual percentages of union voters in each contest will be higher than those listed here. Note that they are more prominent in delegate-rich states like California, New Jersey, and Massachusetts. If Clinton can win unions as strongly as she did in New Hampshire, she will be in good shape.

A similar story can be told about the Catholic population, which also broke decisively for Clinton in New Hampshire. The following chart delineates Super Tuesday states by the percentage of the population that is Catholic.

Catholic.jpg

Note the same features. Delegate-heavy states tend also to have many Catholics. And, once again, Catholics tend to be disproportionately Democratic - so their numbers in the primaries will probably be greater than those represented here. The candidate who wins a plurality of the Catholic vote will be well positioned indeed.

And what about lower income voters? The follow table reviews Super Tuesday states by the average median income for '05 to '06.

Median Income.jpg

As you can see, New Hampshire is wealthier than most Super Tuesday states. So, a candidate who wins voters whose incomes are less than $50,000 per year will probably do better overall than Clinton did in the Granite State.

This indicates to me that if Clinton can replicate the voting coalition she enjoyed in New Hampshire - she will all-but-clinch the nomination on Super Tuesday. I would expect her not just to win most states, but to win them by a wider margin. One of the big reasons is that, while Independents were a feather in Obama's cap in New Hampshire and Iowa - they are barred from voting in Alaska, Arizona, Colorado, Connecticut, Delaware, Kansas, New York, and Oklahoma.

The key word in the last paragraph is "if." We cannot simply assume that Clinton will replicate this coalition. After all, she failed to do it in Iowa. This is actually one of the most interesting dynamics of this primary contest so far.

If we compare the New Hampshire exit polling data with the Iowa entrance polling data, we find a curious divergence. Consider the following table, which reviews the difference between Clinton and Obama among key demographic groups. So, for instance, Clinton won 46% of the female vote in New Hampshire to Obama's 34% - thus yielding the 12% statistic in the corresponding cell. Numbers that favor Obama are in boldface; numbers that favor Clinton are in itallics.

Clinton v Obama.jpg

This, to me, is truly stunning. Clinton won New Hampshire by winning groups that she lost in Iowa. This implies that it is over-simple to categorize Obama as an insurgent candidate who appeals to a wealthy-but-narrow slice of the Democratic electorate (ala Bill Bradley). Obama can win union voters, lower income voters, and women. He did it in Iowa. His victory there demonstrates that he has the capacity to put together a voting coalition composed of traditional Democrats.

Another important question: how will Obama do with African-Americans voters? They are of critical importance in the Super Tuesday states, as the following chart delineates (data courtesy of the 2006 Almanac of American Politics).

African American.jpg

African-American voters have been the most loyal members of the Democratic voting coalition for the last forty years. If Obama can win a strong portion of this vote - he could seriously damage Clinton's prospects.

Can he? We do not know yet. The African-American populations in Iowa and New Hampshire are, as the above chart shows, quite small - too small to infer how the nationwide population will break. Recent polling data suggests he may well be able to bring African-Americans into his coalition. Insider Advantage's recent poll of South Carolina found Obama beating Clinton among African-American voters 48.2% to 36.8%. Survey USA found Obama with an even larger lead. Both of these polls were conducted before the New Hampshire primary - so, it remains to be seen whether these leads are sustainable. [Unfortunately, Rasmussen's post-New Hampshire poll of South Carolina does not offer cross-tabs by race for cheapskates like myself!]

This indicates that South Carolina will be a crucial test of each candidate's viability ahead of February 5th. If Clinton can win the African-American vote - she will be well on her way to the nomination. On the other hand, look out for Obama if he wins that vote. There are some states on Super Tuesday where it could make an important difference.

South Carolina could also be a good test among lower-income white voters. It is a relatively poor state - and so it will be interesting to see how lower-income whites break.

Similarly, Nevada will be an important indicator. Clinton won the union vote in New Hampshire. Obama split it with her in Iowa. Nevada is a relatively unionized state - 14.8% of Nevada workers are unionized. This means that the Nevada returns could give us a clue as to how both candidates will do with union workers on Super Tuesday. The fact that Obama won the endorsement of Nevada's culinary workers' union, as well as the state's service employees union, is a testament to the fact that Obama does have the potential to carry large chunks of the traditional Democratic coalition.

The bottom line: Iowa and New Hampshire diverged in the fullest sense of the word. Not only did Iowa vote Obama and New Hampshire vote Clinton - identical demographic groups broke in opposite directions! This implies that the nomination contest is very much up in the air. The big question is who can put together more of the traditional Democratic voting coalition. Obama did in Iowa. Clinton did in New Hampshire.

There is, of course, an X-factor - momentum. The nomination is a dynamic process. What is interesting from these results is that voters may be slightly indifferent when it comes to these two candidates. That could explain why similar groups voted in different ways - and so the candidate on a "roll" might be the one to win over these groups. Thus, the outcome of February 5th might depend heavily upon the results of Nevada and South Carolina.

If you are interested in digging deeper into the primary battle (and, if you have gotten to the end of this essay, I'll wager you are!) - I encourage you to check out my two-part "Primer on Momentum" if you have not already. It can be found here and here. Unfortunately, it will not give you any hard answers. We just do not know enough about momentum to know when and where it will work. But it should give you a sense of how momentum can operate.

-Jay Cost

How Clinton Won

Hillary Clinton won last night by putting together the voting coalition that has held Democratic frontrunners in good stead for 75 years. Take a look at these numbers - all of which come from CNN's cross-tabulated exit polls. What you'll see is that Hillary Clinton won many elements of the traditional FDR coalition.

-Self-identified Democrats made up 54% of the electorate. She won them, 45% to 34%.

-She won voters without a college degree, 43% to 35%.

-She won voters with incomes less than $50,000, 47% to 32%.

-She won voters over the age of 65, 48% to 32%. She also won voters in their 40s (44% to 33%) and their 50s (39% to 30%).

-She won Catholics, 44% to 27%.

-She won urban voters, 43% to 35%. She won suburban voters, 42% to 35%.

-She won voters from union families, 40% to 31%.

-She won voters who said they have been "falling behind" economically, 43% to 33%.

-She won long-time voters, 38% to 33%.

Obama, on the other hand, had a very different electorate - one that has a bit in common with the insurgent candidacies of Gary Hart and Bill Bradley.

-He won Independents, 41% to 31%.

-He won voters with at least a college degree, 39% to 34%.

-He won voters who make more than $50,000, 40% to 35%.

-He won college age voters, 60% to 22%. He split voters in their late 20s, 35% to 37%. He won voters in their 30s, 43% to 36%.

-He split Protestant voters, 36% to 36%.

-He won rural voters, 39% to 34%.

-He split voters from non-union households, 39% to 38%.

-He won voters who said they were "getting ahead" economically, 48% to 31%.

-He won first time voters, 47% to 37%.

An additional ingredient to Clinton's success was a victory among female voters, 46% to 34%. Obama won male voters, 40% to 29%. But female voters outvoted male voters, 57% to 43%.

As I said, Clinton's is the type of electorate that has delivered Democrats the nomination again and again. These results remind me a great deal of the electorate that delivered Mondale the nomination in 1984 - noting, of course, the irony that Clinton won New Hampshire with this bloc and Mondale did not.

This suggests the model for Clinton moving forward: win by appealing to the traditional Democratic electorate. If she must fight Obama state-by-state, she would do well to reformulate this "Mondale Model" again and again.* This bloc of voters is more sizeable in other states.

Of course, what we do not yet know is whether Obama will be able to win one of the most loyal and potent parts of the traditional Democratic coalition, black voters. My intuition is that black voters will be absolutely critical to the prospects of Obama and Clinton.

Some pundits will probably reference Saturday's debate or Clinton's near-crying moment as reasons she surged late. The exit polling does not back this up. Obama won voters who decided sometime between a month and three days ago. And the two split voters who decided today - 39% to Clinton, 36% to Obama. Clinton dominated among voters who said they decided earlier than a month ago, 48% to 31%.

This supports the idea that Clinton won by mobilizing the traditional Democratic coalition that is demographically inclined to her. You don't just win elections by persuading people you're the best candidate. You win elections by getting those people out to the polls. This appears to be what Clinton did. Accordingly - the implication is that the polls were wrong not because of last-minute shifts. They were wrong because they underestimated Clinton's ability to draw out her base.

Final point. Clinton did something last night that most successful frontrunners have managed to do: use a reliably partisan voting coalition as a counterbalance to an opponent's momentum.

* - The idea of a "Mondale Model" should be taken only as reference to this particular primary strategy - using traditional Democratic groups to hold back an insurgent who appeals to upscale Dems. As Mondale beat Hart, so perhaps Clinton has found a way to beat Obama. I'm not making any arguments about the general election. I certainly am not predicting a 49-state sweep for the GOP should Clinton win the nomination!

-Jay Cost

Clinton's Plan B

New Hampshire Democrats appear to have tipped their hand. They like Obama. He has a lead in all of the major polls, and he is about 7 points up on Clinton in the RCP average.

If Clinton loses New Hampshire - as you, me, and everybody else expects her to - many pundits will declare her candidacy finished. This conclusion will be hasty. Not as hasty as declaring her nomination inevitable, but hasty nonetheless. Clinton still has a viable path to the Democratic nomination.

As I indicated yesterday, she has two real disadvantages that prevent her from reacquiring momentum before Super Tuesday. First, her opposition is not divided. While Edwards might hang around - he probably will not win any states. So, Clinton's loss is Obama's gain. Second, she does not have a geographical firewall. The race is not coming "home" to her.

So, she has lost Iowa. She will probably lose New Hampshire. Many expect her to have trouble in South Carolina and Nevada. Having been stripped of their delegates, Florida and Michigan are non-factors. This means that Obama will probably go undefeated before Super Tuesday. Accordingly, he will be a major force on that day. The latest Gallup poll of the nation's Democrats, which shows a remarkable turnaround for Obama, gives us an indication of just how competitive he will be on February 5th.

What the Clinton campaign needs to do is build a new nomination strategy around this expectation - though it should still work like the dickens to eke out a win in South Carolina or Nevada. But I do not think her campaign can count on these states rescuing her.

It needs a strategy that does two things. First, it appeals to the Super Tuesday states so as to at least split that day's vote with Obama. Second, and what I will talk about today, it borrows from the thinking of the Giuliani campaign: it is all about delegates.

As you probably know, voters do not directly elect nominees. They register their preferences for presidential candidates - either through primaries or caucuses - and those preferences determine how many of a candidate's delegates go to the national convention. The nominee is the person who wins a majority of delegates. Strange indeed - but delegates, just like the conventions they attend, are holdovers from the era of state party dominance.

There are two features of the Democratic nomination process that could help Hillary.

First, Democratic primaries and caucuses allocate delegates proportionally. Candidates win "pledged" delegates based not on whether they win a state - but on how many voters support them. So, for instance, even though Clinton and Edwards lost Iowa, they still won a few delegates.

Second, about 20% of all delegates to the Democratic convention are "super" or "unpledged" delegates. This quirky provision - which does not have a corollary on the Republican side - has its origins in Chicago, 1968. In the wake of that disastrous convention, the DNC formed the McGovern-Fraser Commission to recommend improvements for the nomination process. McGovern-Fraser suggested that the process be opened to rank-and-file Democrats on the principle of "one Democrat, one vote." The reforms contributed to George McGovern (the same McGovern from the commission) winning the nomination in 1972. The party establishment did not like this. So, it added the super delegate provision to serve as a check on the party rank-and-file.

This year, according to the indispensable Green Papers, there will be 798 super delegates at the convention in Denver. They include all elected members of the Democratic National Committee, all current Democratic members of Congress (including non-voting delegates), all sitting Democratic governors, and past party luminaries (e.g. former presidents). Unlike pledged delegates, who are bound to particular candidates, super delegates are free to vote their consciences.

Here is how these rules could help Clinton.

Suppose that Clinton stumbles early, but rebounds later. By the end of the nomination period - she draws even with Obama in the primaries. She wins 45% of the aggregate vote. He wins 45%. Edwards, who in this scenario dropped out some time before the end of the season, wins 10%. That could yield the following count among pledged delegates:

Obama: 1,464 delegates
Clinton: 1,464 delegates
Edwards: 325 delegates

This leaves the 798 super delegates, who can support whomever they choose. Let us suppose, in this scenario, they divvy up the way the Hill reports declared members of Congress have so far split their support between the three major candidates: 62% for Clinton, 25% for Obama, and 13% for Edwards. That would change the delegate count to:

Clinton: 1,967 Delegates
Obama: 1,664 Delegates
Edwards: 420 Delegates

A candidate needs 2,026 delegates to win the nomination. In this scenario, Clinton goes from being tied for first to having a solid lead, and just 58 delegates short of the nomination. If she could persuade about three-fifths of the Edwards' super delegates to back her, she would win.

Now, this is not a prediction about what will happen. It is simply meant to illustrate that the rules of the nomination process give Clinton two advantages.

First, the proportional allocation rule buys Clinton time to get her campaign back on track. This is critically important. Most people assume that February 5th will be the end of the nominating season. Not necessarily. Remember that 44% of all pledged delegates will not be allocated until after Super Tuesday. Clinton could use the proportional allocation rules to keep the delegate count close through February 5th - and then draw even with Obama toward the end of the season. Perhaps as the press starts to examine him with the scrutiny that they give to frontrunners, Democrats will come back to "old rough and ready" Clinton.

Second, a tie between Obama and Clinton would probably be broken in Clinton's favor, thanks to the super delegates. Note that Edwards could diminish the effect of this. If he stays strong - say around 15% of the vote through the primaries - he might have enough delegates to swing the nomination to Obama. However, I doubt that he will stay that strong. He has a hard core of supporters, but like all candidates - many of his voters are strategic. They will abandon him if they realize that he is not viable.

Of course, these advantages alone will not be sufficient to swing the nomination to her. There are caveats. For starters, Clinton needs to tie Obama in the aggregate primary vote to retain her super delegates. They are party elites who would surely recognize the peril of not nominating the candidate who won the most primary votes. She probably could not count on her super delegates to stay with her if the final primary vote is, say, 45% Obama to 35% Clinton. This means that she will have to surge at some point. Relatedly, there are probably a number of super delegates waiting to throw their support behind whoever wins more pledged delegates - Clinton would have to pull even with Obama to pick them up. Also, while delegates make the formal difference, the press, and therefore the electorate, does not see it that way. Clinton will need to win a few important contests soon so that she is still perceived to be viable. This concern may be especially acute, given news reports that she is a little short on resources.

But, the bottom line is this: the rules give Clinton some real advantages. The proportional rule buys Clinton some time to change the dynamic. It will take Obama a while to develop an insurmountable lead among delegates. The super delegate rule gives Clinton the inside track if she does manage to change the dynamic. If she pulls even with him among pledged delegates - the smart money would be on her.

Since George McGovern won his party's nomination - never once has an insurgent candidate defeated an establishment candidate to acquire the nomination. That goes for both parties. In the 220 year history of presidential nominations, insurgents have defeated establishment darlings only a handful of times. A big reason is that the establishment sets the rules. An insurgent candidate riding a wave of grassroots support still has to contend with them. This does not mean that an insurgent cannot beat an insider. It just means that it is hard. Obama will probably win New Hampshire tonight. And he is in a very good position to win his party's nomination. But don't count Clinton out. She's the insider, the candidate of the establishment - and they have a habit of winning.

-Jay Cost

Is 1992 the Model?

With Hillary Clinton's loss in Iowa, and her polling troubles in New Hampshire - her campaign has been spinning that she will soon be the new "Comeback Kid." Like her husband, she will eventually overcome early defeats to win her party's nomination. Bill Clinton himself has noted to several reporters that he did not win his first contest in 1992 until Georgia, which was held a month after Iowa.

So, the implication is that if Bill could lose early contests and bounce back, Hillary could, too.

From a certain perspective, I think this conclusion is indisputable. I do not believe this race is over - and I say that as somebody who predicted that Obama would be a real threat to Hillary a while ago. Here's my bottom line on the Dem race: Clinton has the money, the prestige, and the support to stay in the race through at least Super Tuesday, even if she loses all of the early contests. She also has, at least according to the latest national polls, much of the traditional voting coalition that has won her party's nomination in year's past. And remember - most Democratic primaries allocate delegates to the national convention proportionally, which means that losers still win delegates. So, Clinton could stay a close second through most of the season, and surge late to win the nomination. Of course, losses in Iowa and New Hampshire would seriously damage her campaign. No candidate who has won both Iowa and New Hampshire has ever failed to win his party's nomination in the modern era.* However, as I have argued many times, history is a limited guide for us when it comes to party nominations. Hillary remains a candidate with real strengths - and she should not be underestimated.

My quibble here is using 1992 as a model for Hillary, which is exactly what her husband and other surrogates have floated. There are two important problems.

Before we get into them, let's review the calendar of events from the 1992 Democratic primaries. As you might remember, Bill was the big polling leader heading into the fourth quarter of 1991. However, the "character issue" seriously damaged his numbers, threw the race into the air, and induced Bill and Hillary to appear on 60 Minutes after the Super Bowl. So, here is how the events went down subsequent to that:

February 10, 1992: Iowa senator Tom Harkin wins his state's caucus. No surprise here. No candidate was challenging him.

February 18, 1992: Former Massachusetts senator Paul Tsongas wins New Hampshire. Clinton finishes second and brands himself the "comeback kid."

February 25, 1992: Nebraska Senator Bob Kerrey wins South Dakota.

March 3, 1992: Mini-Super Tuesday. Clinton wins Georgia. Former California governor Jerry Brown wins Colorado. Tsongas wins Maryland. Harkin wins the Minnesota and Idaho caucuses.

March 7, 1992: Clinton wins South Carolina.

March 10, 1992: Super Tuesday. Clinton wins six southern states: Florida, Louisiana, Mississippi, Oklahoma, Tennessee, Texas. Tsongas wins Rhode Island and Massachusetts.

March 17, 1992: Clinton essentially ends the contest with a win in Illinois (though Brown would briefly reemerge with a victory in Connecticut one week later).

So, what did it for Clinton? The South! The South dominated Super Tuesday, which was the reason why it was created. Super Tuesday was designed by southern Democrats after Mondale's nomination in 1984. They wanted to use their weight to nominate a more moderate candidate who would better reflect their interests. The plan backfired in 1988, as Al Gore and Jesse Jackson effectively neutralized one another, and Michael Dukakis won the nomination. But the plan succeeded in 1992 - as Bill Clinton lost seven of the first nine contests, but still won the nomination.

This points out a critical difference between Bill and Hillary. Bill won the nomination when the battle came to his home turf. The South was Clinton's firewall in 1992. It was going to get behind Bill almost regardless of where he finished in the previous contests. Hillary has no firewall that is based upon regional affinity. Of course, she is currently strong in many of this year's Super Tuesday states - California, New Jersey, Connecticut, etc. However, she is not a favorite daughter. These states are not nearly as dependable for her as Tennessee, Mississippi, and South Carolina were for Bill.

Another critical difference can be noted when we look at the results of the early contests. Of the contests prior to Super Tuesday, they were split five ways: Harkin, Kerrey, Tsongas, Brown, and Clinton all won at least one contest before March 10. There was no significant consolidation of the race because the early states kept disagreeing with one another. This helped Bill. He could lose and lose and lose because no single opponent won and won and won. Hillary does not enjoy this kind of advantage. The 2008 Democratic contest is between Clinton and Obama. So, if she loses seven of the first eight contests contests, Obama wins seven of the first eight. This would create a dramatically different dynamic than in 1992.

Thus, we see here two of Clinton's relative weaknesses as a frontrunner. She has no firewall that stems from her geographical roots. She also does not have the luxury of multiple opponents.

If the 1992 Democratic contest has a parallel with this year, it is on the Republican side. And Rudy Giuliani may be Bill Clinton. Like Clinton in 1992, Giuliani's field has multiple viable candidates. Also like Clinton, he has a natural constituency upon which he might draw. Obviously, he is not from Florida - but a lot of Floridians are former New Yorkers. This gives him a real advantage in the Sunshine State. Also, several of Rudy's best states are holding their primaries on Super Tuesday: New York, New Jersey, Connecticut, Massachusetts.

Of course, the parallels do not hold terribly well. First, Clinton placed a strong second in New Hampshire. Giuliani is unlikely to do that. This could damage him - especially if he drops behind Huckabee or even Paul. Second, this year's Super Tuesday includes seventeen states where Giuliani cannot be said to have a natural constituency. There are a lot of moderate states, for sure, but we are talking about regional affinity here. Third, you could argue that the South is much more loyal to its sons than the northeast is. I cannot imagine Louisiana voting for anybody but Bill Clinton in 1992. I could definitely see Connecticut and Massachusetts going for McCain - just like they did in 2000. All in all, while 1992 should give Rudy some comfort, it is not a tight fit with his current position.

As for Hillary Clinton - she is far from finished, even if she loses New Hampshire. But her husband's come-from-behind win in 1992 cannot be her model. Her husband had advantages that she simply lacks.

*. Update, 3:30 PM: A reader points out that Edmund Muskie won New Hampshire in 1972, and was the highest vote-getter in the Iowa caucus. In light of this observation, I have amended the above sentence to include the phrase "in the modern era." The modern era in the presidential nominating contest is usually said to have begun in 1976 - which is the first year that primaries and caucuses dominated both party contests. I apologize for the vagueness in my original piece.

-Jay Cost

On the Iowa Results

A few thoughts on the Iowa results:

(1) My initial impression, which I am sure you share, is that this is being chalked up as huge wins for Obama and Huckabee, huge losses for Clinton and Romney. This interpretation could be a big deal. Remember that is exactly what it is. The press is not interpreting this as "Clinton ties Obama among Democrats in entrance poll" or "Mormon Romney finishes strong second in evangelical Iowa." This matters. Watch how the press continues to interpret these results over the next few days. It is the source of information for persuadable voters in New Hampshire. Perceptive observers of the horse race should move beyond the nouns, and start looking at the adjectives and adverbs. They make a big difference.

(2) It's gut-check time for the Clinton campaign. Watching Chris Matthews et al. on MSNBC last night - they were talking like it was all-but-over for Clinton. No way. New Hampshire is the early state that has the biggest impact. Not Iowa. Iowa has a habit of picking losers. It is easy for media types to forget that because in its most recent outing, 2004, it single-handedly determined the winner. But historically, Iowa does not make much of a big ripple nationwide. The big question: will Iowa move New Hampshire? Obama needs it to. He is in second there right now. We don't have an answer yet - and history provides a mixed message. Sometimes Iowa does move New Hampshire. Sometimes it doesn't.

(3) Huckabee is going to get a big surge in positive press - but he has two big problems. First, everybody is gong nuts about Huckabee tonight, but on Tuesday they'll be going nuts about McCain. That limits the effect of this win. Second, the next contests are coming much too soon for this media attention to yield dollars, and the dollars to yield organizations that can help him on February 5. Huckabee wants to run a campaign reminiscent of Carter '76 (in more ways than one), but Carter had one thing Huckabee does not: a schedule that is spread out.

(4) Tonight was bad for Romney. Really bad. He lost by a lot. He lost by more than anybody expected. He lost after having led for a year. He lost after a monumental effort. This loss was bigger than Clinton's. He is not his party's frontrunner. He cannot afford to lose a state he tried so hard to win. Worse for Romney - McCain had already surged ahead of him in New Hampshire prior to tonight's loss. This will not disrupt that dynamic. I would not go so far as John Ellis - but I will say that he is in really bad shape. His future in this race is bleak.

(5) Yesterday was the first official day of Rudy going "dark." He has no real plans to reemerge until January 29th. Is this a viable strategy? Honestly - I think it may be. If he wins Florida, he surges back into the race. I think Florida could hold for him, though I have real doubts. A lot of that depends on who is still in the race at that point. At the least, I think this is the best strategy that a candidate like Giuliani could pursue. A Giuliani nomination was always going to be a close call. It was always going to be a victory over a divided party. Giuliani should be happy that Huckabee has damaged Romney because Romney has money and Huckabee does not. Rudy wants February 5th to come down to Rudy versus the anti-Rudy candidate - and the less well-funded the anti-Rudy candidate, the better. But a McCain victory in New Hampshire should make Giuliani nervous - especially if it is as big as Huckabee's in Iowa. A three-way race between McCain, Rudy, and Huckabee would be harder for Rudy because he and McCain occupy much of the same ideological space. Rudy would probably like to see Huckabee hold the line in South Carolina.

(6) Fred Thompson is finished. Absolutely, positively finished. The reason? He has no more money - which is the reason all losing candidates drop out. And this defeat tonight is not going to get him any cash. The big Thompson question on my mind: if he drops out and endorses McCain, does that swing the 14% or so of South Carolinians who currently support Thompson? [Update, 2:30 AM: As of this evening, Thompson promises to forge ahead.]

(7) The same goes for Romney. If Romney fades from view after Tuesday - where do the Romney supporters go? Who are their second choices? Are they social conservatives who will support Huckabee, or are they voters who want hyper-competent executive management and a tough stance in the global war on terror? I have no idea. Nobody does. The polls don't tell us.

(8) Edwards is also done, in my opinion. He can't have much money left. He's in distant third everywhere in the nation. And nobody will be talking about him anymore. I expect him to linger - but it is clear that he is not the anti-Hillary candidate on the Democratic side. The big question: where does the 10-15% of the Democratic electorate who supports Edwards go?

(9) Ron Paul remains a non-factor despite his money and his strong showing. The reasons are two-fold. First, he cannot win the nomination. Second, he does not hurt anybody. My guess is that he is bringing new voters into the process. Good for him - but he is not a player in his party's nomination contest. (Update, 2:11 AM: A few readers have suggested that he hurts McCain. Maybe he will in New Hampshire. Both probably appeal to independents. But McCain is not up against George W. Bush in New Hampshire this time around. He's up against Romney - and he has a lead. While Paul may take some votes away from McCain, at this point it does not appear that he diminishes McCain's chance of winning New Hampshire.)

-Jay Cost

Should Clinton Have Skipped Iowa?

On New Year's Day, Roger Simon asked:

Should Hillary Clinton have skipped Iowa?

If she loses the caucus here Thursday, will her campaign wish it had listened to the advice it got last May to take a hike on the Hawkeye state?

Back then, Clinton's deputy campaign manager, Mike Henry, wrote a 1,500-word internal memo saying Iowa was not worth the effort.

"My recommendation is to pull completely out of Iowa and spend the money and Senator Clinton's time on other states," Henry wrote.

"If she walks away from Iowa she will devalue Iowa -- our consistently weakest state."

Henry's advice was never accepted.

After the memo was leaked to the press, the Clinton campaign publicly repudiated the memo and said it would compete fiercely in Iowa, which it has.

Others have been asking this question in the wake of Simon's column.

I would argue that the answer is an unequivocal "no." Clinton's campaign was wise to have stuck Iowa out. I believe that it would be better off losing Iowa today having run a full campaign than it would have been if it had avoided the Hawkeye State altogether.

It's a matter of costs and benefits. Skipping Iowa would have provided a benefit, for sure. She would not have been as "diminished" if she had left than if she had stayed, fought, and lost. But I think this benefit is far outweighed by the costs.

First, frontrunners like Hillary Clinton do not just skip states. They compete - unless one of their opponents is from the state. They have the money - and they want to be the party's nominee, so they presumably have the message. Skipping a state that you could compete in gives a horrible impression. It rightly inclines all of the partisans and media types who have made you the frontrunner to reconsider their decisions. Any reconsideration would have hurt Clinton.

Second, the media circus that descended upon the supposedly worthless Ames straw poll should put the lie to the idea that Clinton's avoidance of Iowa would have devalued it. Journalists are desperate for news. They would have turned Iowa into a story as long as Obama and Edwards were competitive with one another - and they can take a broad definition of "competitive" when they really want a story. Taking Clinton out of Iowa might have reduced the number of credentialed people in Iowa right now - but not by all that much. The caucus results would still get basically all of the attention that they will receive - only in this scenario, all of the stories would remind voters (five days before the New Hampshire primary, mind you) that Clinton skipped Iowa because she thought she would lose.

Third, Iowa has been known to give candidates momentum heading into New Hampshire - as happened in 1976 and 2004. Ceding Iowa would dramatically increase the chances that Obama comes to the Granite State with a head of steam. With Clinton out of Iowa, all he would have to do is beat Edwards - and he could bask in the glow of stories praising him as the "alternative to Hillary" for the next five days. This would have been very bad - because New Hampshire is the critical early state in most cycles. Now, of course, by skipping Iowa - Clinton could have husbanded her resources and forced Obama to spend his cash just to beat Edwards. There would be some benefit there, but not as much as one might think - after all, Obama has lots of money. He would still have plenty to compete.

Fourth, and most important of all, she could very well win Iowa! It seems clear to me that the differences in the polling data we have seen is dependent upon the turnout models that the pollsters are using. Under certain assumptions, Clinton leads in the polling. Under other assumptions, she does not. This implies that if those favorable assumptions hold today, Clinton will win the caucus. This would be a huge boost to her candidacy. The shot at winning the first contest is worth the risk of competing.

As I said, I would agree that skipping Iowa might have mitigated some of the diminution that she would suffer upon losing Iowa to one of her challengers. But, it would have raised questions about her frontrunner status, it would not have diminished the importance of the caucus, it would have given Obama a leg up heading into New Hampshire, and it would have taken away her shot at ending this contest early. Combined, these costs are greater than this benefit.

And, here's the most important point: virtually every frontrunner who goes on to win the nomination suffers at least one setback in the campaign. Many of them suffer more than one. What makes them strong candidates is that they move forward and overcome the initial defeat. This is what separates Howard Dean from most nominees. If history is any guide, Clinton should expect to lose an important one at some point in this cycle. If she has what it takes to be the nominee, she'll bounce back - just like Ronald Reagan, Walter Mondale, Michael Dukakis, George H.W. Bush, Bill Clinton, Bob Dole, and George W. Bush.

Clinton was wise to stay and fight. It was her best option.

-Jay Cost

An Iowa Push?

Adam Nagourney asked an interesting question in yesterday's New York Times.

"What if Iowa settles nothing for Democrats?"

As of now, the question has no answer. It is a device for Nagourney to make the following, worthwhile observation:

What if at the end of Thursday, the three leading Democrats -- former Senator John Edwards and Senators Hillary Rodham Clinton and Barack Obama -- are separated by a percentage point or two, leaving no one with the clear right of delivering a victory speech (or the burden of conceding)? A number of polls going into the final days have suggested that after all of this, the Democratic caucus on Thursday night could end up more or less a tie.

In truth, amid all the endless permutations of outcomes that are being discussed -- can Mrs. Clinton, the putative front-runner, survive a third-place finish, or Mr. Edwards a second-place one? -- aides are beginning to grapple with the frustrating possibility that all the time, money and political skill invested here might prove to be for naught when it comes to identifying the candidate to beat in the primaries and winnowing the top tier.

This is a great point. For months, I have asserted that an Iowa victory is a necessary condition for Obama to be nominated, but a sufficient condition for Clinton. I always assumed - unthinkingly, in retrospect - that somebody would win Iowa. But Nagourney articulates a question that a lot of us have probably been wondering: what happens if nobody really wins? How will the remainder of the race be influenced?

This question is actually all about momentum, a subject I have recently covered in depth (See Part 1 and Part 2 of "A Primer on Momentum"). I'd like to draw upon those considerations to make a few comments on Nagourney's interesting article.

First, the prospect of Iowa not influencing the race should always have been considered possible, if not likely. As I noted in Part 2, Iowa generally has not been influential in either party's nomination battle. Occasionally, it has affected New Hampshire - as it clearly did in 2004 - but this is a relatively rare phenomenon in the modern era.

Second, we must emphasize the media's role in determining who "wins" Iowa, or any other state. Nagourney punts on this subject later on when he argues that Bill Clinton, "proclaimed himself the winner after coming in second in New Hampshire." Well - sure Clinton proclaimed himself the winner. But the only way that proclamation stuck was that the media agreed with it - which goes to show that, in situations like this, the perception of the media is as good as reality. Iowa will be a push if and only if the media interprets it as such.

This is why I took Nagourney's article to be especially valuable. He raises some good points, for sure. But it is also important that he raises them. Nagourney is one of the nation's top political journalists. If he is anticipating that Iowa might be a push - that is a sign that he, and the rest of his fellow journalists, could ultimately interpret it as such. This is why I would disagree with Tom's comment yesterday that, "the idea of a "push" is antithetical to how the media operates." If Nagourney is willing to consider the idea - the press might eventually embrace it.

Third, the fact that we cannot answer Nagourney's question is indicative of a point I made in Part 1: we simply do not know that much about how momentum works. Our dataset is just too limited to infer what an Iowa push will mean this year.

Nevertheless, we can make use of our knowledge about momentum to frame the question a little more precisely than Nagourney does. We know that the effect of momentum begins with the individual voter. We also know that the next contest - and thus the one to be affected by Iowa directly - is New Hampshire. Finally, we know that the 800-pound gorilla in the room is Hillary Clinton. Ultimately, what will be of decisive importance is how voters view her. So, our precise question is: how will an Iowa push affect New Hampshire voters' views of Hillary Clinton's candidacy?

You can make a convincing case that an Iowa push hurts Clinton via several of the five types of individual-based momentum discussed in Part 1; but, you can make a similarly convincing case the other way. That's exactly the point. We do not know enough about momentum to draw a firm conclusion. Nor, for that matter, do we know enough about what New Hampshire voters are thinking at this moment. While reams of polling data have poured forth from the Granite State - most of it probes the basic horse race questions that interest the media. Hardly any of it touches upon voter psychology - and so hardly any of it helps us gauge how New Hampshire voters will react to a Hawkeye push.

At the least, though, we know what to watch for. We should watch for how the media interprets the Iowa results. If the media calls it a push, it's a push. If it doesn't, it isn't a push. Our personal interpretations will not matter - the collective interpretation of journalists and pundits will. We should also watch for any trends that appear in the New Hampshire polls. We should have time for a round or two of post-Iowa New Hampshire polls. They will be our first and best clues on how New Hampshire has reacted to the results of the caucus.

-Jay Cost