No, no, no. The Republican presidential nomination really is not over despite the parade of rather craven endorsements now flowing Mitt Romney’s way.
Before I go any further, a few words of explanation. Please bear with me. Part One of this column will explain why I am writing this right now. Part Two, the meat of it, will explain the actual delegate arithmetic in play, while also providing part of a tutorial on caucus/convention systems. Part Three will “game out” the remaining contests. Part Four will explain why a contested convention could be very much to the benefit of the Republican Party and its presidential candidate.
It must first be acknowledged that I am not at all an objective observer. In this race, I have become an outspoken Rick Santorum partisan — so much so that I wrote a month ago that, with a few stated exceptions, I would no longer columnize (to coin a word) on this race. Indeed, I have refrained from commentary on the overall, relative merits of the candidates. But now I will make a slight exception — not as candidate advocacy, although I will be accused of it, but in the realm of explanatory journalism. As I do so, please bear my biases in mind, and judge accordingly.
I make this exception because, before I went full-time into journalism, I worked on lots of campaigns and particularly took pride in my work in caucuses and conventions, where I helped achieve some significant triumphs, at least for the scale of event involved. (One case in point, relevant to my coming argument, was when the number of David Duke supporters who showed up at one particular legislative district caucus was greater than the combined supporters for two other candidates and for an “uncommitted” slate — but, by using smart tactics and good discipline, the non-Duke bloc, thank goodness, secured four of that small district’s five delegate slots to the state convention. This helps make an important point we shall see later, which is how pluralities or even intelligent minorities of voters can secure majorities of delegates from a caucus.)
It therefore galls me to see media accounts, and delegate counts, that misrepresent the possible or likely results of district and state confabs, which in truth can produce outcomes remarkably different from primaries. Because much of the “race is over” narrative relies on these potentially mistaken delegate counts — and because about half of all the nation’s Republican voters still haven’t had a chance to weigh in on this contest, and will effectively be disenfranchised if the battle is concluded prematurely — it is extremely important for both pundits and voters to understand where the narrative may be flawed.
The Santorum campaign has asserted that the delegate race is notably closer than the punditocracy has indicated. Their contentions largely have been dismissed by that same punditocracy, with some analysts describing the Santorum claims as inhabiting “fantasyland.”
Here’s why the truth is probably somewhere in between.
To understand why, one must understand that the differing assertions almost all involve caucus/convention states, and must understand that those states tend not to award delegates proportionally. Instead, they encourage results that approach “winner-take-all” decisiveness.
As a hypothetical, imagine a district caucus in which only two candidates are competing to choose, perhaps, five delegates to a state convention. Now imagine that 199 voters participate in the caucus. Imagine that exactly 100 of those 199 are firmly committed to one of the two candidates. If those 100 are disciplined, and can agree among themselves to let no more than five of their number run for the delegate spots, then they will win all five of those delegate spots regardless of what the opponents do. Why? Because all five of Candidate A’s delegate designees would receive exactly 100 votes, while none of Candidate B’s delegate designees could receive more than 99 votes. An almost perfectly evenly split caucus (100-99) would therefore produce a unanimous delegation to the state convention.
The same dynamic would apply at the state convention. If one candidate enjoys bare-majority support at the state convention (after all the delegations from each of the district caucuses were assembled), that candidate could use a disciplined organization to sweep every single one of the delegate spots for the national convention. The media may report caucus results statewide as being almost exactly 50-50, and its delegate estimates would therefore assume an almost even split — but those estimates would be horribly wrong.
Granted, caucuses and conventions rarely work out in quite so clear-cut a fashion. Presidential campaigns rarely can be so certain of the candidate preferences of caucus voters or convention delegates. And, of course, if those delegates are not bound by law to stick with their originally stated preference, they can always change their minds. Nonetheless, the principle still applies: Conventions can magnify small advantages.
(Also, of course, if the 100 for Candidate A are not disciplined, but the 99 for Candidate B know what they are doing, then the 99 can actually outvote the 100. How? Because in a caucus, people vote directly for delegates, not for a candidate. If some of the 100 vote for delegates not pledged to their candidate of choice, then those votes are accordingly diluted. This actually can happen quite easily, if, for instance, some well-respected potential delegates earn votes from some of the 100 erstwhile supporters who are confused by the system and vote for delegates based on the delegates’ perceived merits rather than based on the candidate to whom those delegates may be committed.)
In a slightly different way, the same principle applicable to bare majorities applies to bare pluralities, too, when more than two candidates are competing. Even more importantly, coalitions between candidates can overcome a plurality enjoyed by a third.
Again, a hypothetical: Imagine a state where Romney secured 40 percent of delegates to the state convention, while Santorum secured 30 percent, Newt Gingrich 22 percent, and Ron Paul eight percent. If the Santorum team and the Gingrich team joined forces, either because of philosophical affinity or even just for the tactical goal of slowing down Romney’s march, they would together be in the same position as Candidate A’s majority described in the first example. They could agree to divvy up the national-convention delegates among their respective camps, with, say, two-thirds of the delegates going to Santorum and one-third going to Gingrich. Even with a plurality of state-convention delegates, Romney would enjoy the allegiance of not a single one of that state’s delegates to the national convention in Tampa, because not a single one of his delegates would win a majority.
This is, by the way, what the McCain camp, masterminded on site by former Louisiana Gov. Buddy Roemer, did with the Huckabee camp at the 2008 West Virginia convention, thus blocking Romney from what would have been a key win in his challenge to McCain.
Other permutations and confusions are possible as well, but the practical outcome remains the same: What matters is not the original percentage of voters; what matters is which camp or coalition can first find a way to a bare majority.
And there is nothing unfair about this. It’s very similar to a state primary system for, say, governor, in which a runoff primary is required if nobody in a multi-candidate race earns 50 percent in the first primary. Backers of the third-place candidate easily can, and often do, swing behind the second place candidate in the runoff, putting him over the top and denying the nomination from the first-primary plurality winner. The same principle applies here.
In real life, take Washington State as an example. Based on raw voter preferences, the Real Clear Politics estimate is that Romney will take 25 delegates, Paul 8, and Santorum 7. Santorum’s team has claimed in public that this count will prove to be wildly off. Well, if the local supporters of Paul, Santorum, and Gingrich tend to work together to block Romney, the Santorum claim could easily prove accurate, with Paul and Santorum each gaining another four or so delegates, all at Romney’s expense.
In a conference call a couple of weeks ago, Santorum spokesmen said they expected to do better than media estimates not just in Washington, but in Iowa, Minnesota, Missouri, Colorado, and perhaps Alaska as well. Because the national dynamic so far has provided practical incentives for the other three candidates to make common cause against Romney — and because Santorum did well enough in those states to be in the driver’s seat within those coalitions — at least some of his campaign’s assertions have the ring of truth. Granted, campaigns looking for silver linings may convince themselves they are doing better than they actually are. But there is every reason to believe that even if the Santorum team is too optimistic, the media counts probably do overstate Romney’s actual delegate lead to a certain extent — specifically because of the dynamic described above.
The Santorum team also plans to challenge the winner-take-all allocations of Florida’s 50-member delegation and Arizona’s 40. There is no doubt, at all, that national party rules dictate proportional allocation for both states rather than a clean sweep for Romney. That doesn’t mean a challenge to those delegations will actually succeed, but there is good reason to think it has a decent chance. Six-term Virginia National Committeeman Morton Blackwell, a GOP rules guru for 40 years, explained in an essay for Red State exactly why such a challenge will have at least a decent shot, and how it would work. “Any likely remedy,” Blackwell wrote, “would reduce Gov. Romney’s delegate vote total and increase the delegate votes for other candidates. In a closely divided convention, that could decide who wins the presidential nomination.”
Some back-of-envelope calculations of my own show that Romney could “lose” a combined 49 delegates from current allocations in those two states, while Santorum could pick up about 28 of those (with Paul and Gingrich each gaining some as well). That would be a significant shift.
Before the Louisiana primary, I “gamed out” the results for the remainder of the presidential nomination contests, with close attention to the precise delegate-allocation rules in each individual state. I did so with two priorities in mind: First, every assumption I made about future results had to be well within reason; second, within that realm of indisputable reasonableness, I resolved most predictive doubts slightly (but only slightly) in Santorum’s favor, as if he were to continue slightly outperforming polling and anecdotal indicators, just as he has in almost every state so far.
For instance, most pre-Louisiana polls indicated Santorum might get seven or eight delegates from its primary to Romney’s six (or maybe as few as five). When I made my assumptions ahead of time, I gave Santorum 9 to Romney’s 5. As it was, I slightly underestimated Santorum’s final margin; Real Clear Politics says he won 10 delegates, Romney five.
When assessing the remainder of the race, I tried to be just as careful. Meanwhile, I assumed that the caucus states mentioned above would indeed, in the end, give Santorum more delegates, and Romney fewer, than Real Clear Politics estimates — but nowhere near to the degree that Santorum’s own camp seems to expect. In short, again, I tried to stay well within reason, but to give a realistically based shading, just slightly, in Santorum’s favor — based not on my wishes, but on Santorum’s already-established habit of outperforming polls.
I did nothing to “massage” the numbers, but let them pile up on their own without me trying to figure how closely they would move Romney to the 1,144 majority he needs for a first-ballot victory.
What I found was that Romney fell just short. Close, but not quite there.
To avoid the “fallacy of misplaced concreteness,” which is what happens when somebody reaches too precise a conclusion despite basing it on a series of contingencies, I will not lay out, state by state, how many delegates I “awarded” to each candidate, nor will I say precisely how close Romney came. The point is not that I have a perfect crystal ball; the point is that, based on long experience and success in this political-prediction business, I think it likely that my errors to Romney’s detriment, state by state, will approximately balance out my errors in his favor — as in Louisiana — meaning that, if my assumptions are right, he could well fall just short.
Now this isn’t to say a first-ballot failure precludes Romney from winning on the second ballot. It just means that the arithmetic gives credence to the idea that he can indeed be blocked.
For an example of how I gamed it out, looking forward after tomorrow’s primaries in Wisconsin and Maryland, I assumed Santorum would win at least pluralities in the following states (through May), all of which are quite reasonable assumptions: Pennsylvania, Delaware (barely), North Carolina, Indiana, Kentucky, West Virginia, Arkansas, Texas, and Nebraska.
Meanwhile, I did not assume a successful challenge to the Arizona and Florida delegations. The Santorum team has said publicly that they can succeed even if they don’t win Wisconsin, but they are counting on winning the rules challenge in the two sun states. But to be on the safe side, I am leaving Arizona and Florida alone for now. Even then, again, the arithmetic shows Romney can be blocked.
Also to show restraint, I did not award wins to Santorum in other states where he is running relatively close, such as Oregon.
Again, this is not a wish list, but an honest assessment. Romney still isn’t a shoo-in; Santorum still has a real chance to force a second ballot at the national convention — and then, all bets would be off.
It must be acknowledged that there are a lot of people, “establishment moderates” and conservatives alike, who for some strange reason are scared stiff by the idea of a seriously contested convention. For decades, I have thought their fears are groundless, and that a contested convention could instead provide a huge boost to the party in which it happens. (Remember that it was only a week or two before the Democratic convention four years ago that Hillary Clinton finally had to admit she had lost to Barack Obama; somehow, that example of a close and fiercely contested nomination battle did not harm their party’s ability to win in the fall against a candidate, John McCain, who had several extra months to prepare and to consolidate Republicans behind him.)
To be clear, conservatives should not wish for a brokered convention. Technically, a brokered convention is one in which, yes, power brokers pull strings behind closed doors and shift entire blocs of delegates with them, as if the delegates are sheeple. This scenario, of course, would cause a public-relations nightmare, with the media going nuts sliming Republicans for resorting to tawdry deals from proverbial smoke-filled rooms.
But that’s not likely to happen. Party rules don’t lend themselves to those occurrences anymore. A lot of states at one point had something called the “unit rule,” which stated that a state’s delegate votes would be cast unanimously for whoever has a bare majority within the state. The unit rule is now a thing of the past. Also, huge numbers of delegates once ran as openly “uncommitted,” rather than even being informally pledged to a particular candidate. Again, those uncommitted delegates tended to be deliverable en masse by powerful party big-wigs. But far fewer delegates these days are now elected while openly “uncommitted” as part of a deliberate strategy to reserve power for the big-wigs.
Instead of a brokered convention, what is needed — and what is likely to happen if no candidate enters a convention with majority support — is a contested, open convention. This would be a good thing for the party.
First, the drama would be riveting. The convention would actually mean something rather than being a mere propaganda vehicle — a vehicle the public increasingly sees through, and increasingly tunes out. The public would watch, rapt, just as they watch so many types of “reality TV,” except that in this case, unlike in artificial scenarios such as in Survivor, the “reality” would actually be real. It would actually mean something, not just for those in the competition, but for the viewers, too. The fate of their country would hang in the balance; viewers would have a real stake in the fight.
Second, what the public would see would be anything but a bunch of politicians trading favors. Instead, what would soon become clear would be what has been lost in these last three decades of uncontested conventions: namely, delegates are mostly ordinary people, volunteers rather than party officials, who are in it not for perks or privileges but because they care deeply about their country. They are housewives, small-business owners, professional women, or retirees worried about their grandchildren. The media would be unable to draw nasty caricatures of delegates as a group, as they do at every quadrennial Republican confab; instead, when each individual’s vote actually makes a difference, the delegates would be interviewed exactly as the individuals they are — and viewers at home would identify with their dilemmas.
The public would see “ordinary Americans” making the momentous decision about who should be the nominee seeking the post powerful position in the world. The resulting impression would be a very good one for the party and the eventual nominee. Not only that, but the whole scenario would re-educate a now-cynical public in the idea that participation in civic life really matters, and that this is still a political system of and by the people. Result: A reinvigoration of the civic order, one that would especially inspire those who empathize with the conservative vision of the Republican Party.
The nominee who emerges from a contested convention could thus do so with a huge surge of momentum to take the fight to Barack Obama in the fall.
Therefore, if the voters of Wisconsin and following states decide to fight back against the establishmentarians and vice-presidential wannabes who tell them their votes shouldn’t matter, the very reasonable chance still exists for Mitt Romney not to sail to a first-ballot convention victory. Even if he does ultimately win on a subsequent ballot — a far from certain result — he might benefit as much as anybody from the magic a contested convention could produce.
The establishment is dead wrong to try to end this contest prematurely. The voters deserve to have their say.