Marco Rubio’s foolhardy persistence in the Republican presidential primary, along with long-term structural biases in the primary process, have made the path to nominating a conservative Republican more difficult than it ought to be.
The good news, however, is that Ted Cruz can still deny Donald Trump a majority of the Republican delegates and win a plurality himself, even if Trump earns what should be an expected five-state sweep next Tuesday night.
As Robert Eno argued last week, Cruz would very likely have (substantially) more delegates than Trump had Rubio and John Kasich chosen to drop out after their poor showings on Super Tuesday (March 1). Moreover, as Sean Davis argued, Rubio’s pursuit of a long-shot victory in Florida could cost Cruz many delegates in the other states voting that day—and perhaps Kasich a winner-takes-all victory in Ohio, despite Rubio’s endorsement of Kasich there.
The Two Simultaneous GOP Primaries
But the consequences of Rubio’s decision to remain in the race are even greater when one takes into account the most overlooked predictor of electoral outcomes so far: whether a caucus or primary is open (non-Republicans are eligible to vote) or closed (only Republican voters).
Thus far, there have been 14 open and 10 closed votes. If we exclude Texas (an open primary), where Cruz’s “favorite son” status seems to have overwhelmed all other factors, the differences in the results are dramatic:
After next Tuesday, there are 19 primaries left, 13 of them closed, an important advantage for Cruz in a one-on-one contest with Trump. But on Tuesday itself, there are four open primaries and only one closed: Florida. Rubio, then, hurts Cruz in the March 15 state where he would otherwise be best positioned to beat Trump, while promising nothing more for Rubio than a stay of political execution, even if he succeeds in his longshot bid for victory.
There are, in essence, two different primaries being run simultaneously. In the states where Republicans choose their own nominee, Cruz has a small advantage over Trump, with Rubio very far behind. In the states where Independents and, in some cases, Democrats are eligible to vote in the Republican presidential primary, Trump is trouncing the field, with Cruz and Rubio in a tight contest for a distant second place (although, by including Texas, Cruz gains a substantial delegate edge over Rubio).
The Third Factor: Romney Votes
How crucial is this? Consider the broader context. Trump clearly hopes to end the campaign next Tuesday. Wins over Kasich in Ohio and Rubio in Florida would certainly knock each of them out of the race. The perception of Cruz’s performance in Illinois, Missouri, and North Carolina, then, will likely go a long way to determining whether there is a real campaign going forward or Cruz appears to be only playing the spoiler—the difference, among other things, between Trump being forced to debate Cruz one-on-one and having a plausible pretext for refusing.
Unfortunately, one may predict with some confidence that Cruz will have “disappointing” results in those three states. A simple analysis of the battle between Trump and Cruz to date shows that half of the variability in their state-by-state performance against each other can be explained by looking at whether a vote is open or closed and just one more variable: the percentage of votes for Mitt Romney in the 2012 general election.
The better Romney did against President Obama, the better, on average, Cruz does against Trump. In sum: the more Republicans in the primary and general electorate in a state, the more likely Cruz will beat Trump. Apply this model to the three “neutral” March 15 states and you find Cruz is unlikely to beat Trump in any of them:
Polling data suggests Cruz will beat the model’s projections in the three neutral states on Tuesday. This is not entirely surprising because, since Ben Carson suspended his campaign, Cruz has exceeded his projected vote percentage in six of eight states, an indication that Cruz is doing better among those who would otherwise be Carson voters than Trump does. Nevertheless, anything close to an even showing against Trump in these three states would be an impressive achievement for Cruz.
In this case, a 50 percent projected share means Cruz would be expected to get the same number of votes as Trump, so anything less than 50 percent means a loss in a two-way race. Looking ahead, there are only three (of 19) states where Cruz can be expected to do worse than the best of these, Missouri. By contrast, were it not for Rubio’s “favorite son” status in Florida, Cruz would be expected to run even with Trump there, potentially winning enough delegates in that winner-takes-all state to erase the entire gap to date between the two candidates.
Time to Reconsider Open Primaries
Of course, primary elections are not won with unexpectedly close second-place finishes (ask Bernie Sanders). But neither should an ill-informed commentariat be allowed to pronounce the contest over when a five-state Trump sweep is perfectly predictable based on the results so far—and no special evidence that Trump would, in fact, win a one-on-one race against Cruz.
On the contrary, if one assumes that, after March 15, Cruz will earn the lion’s share of support from would-be Rubio and Kasich voters, as recent exit polling data indicates, our simple model suggests Cruz would be well positioned to prevent Trump from earning a majority of GOP delegates and could still enter the Republican convention with a plurality of delegates himself, as the likely winner in at least 12 of the 19 remaining state contests.
The point of all this is not to make overly precise predictions about the future but to suggest:
- That, within the bounds of reasonable assumptions, Cruz is already doing what is necessary to (at least) deny Trump a primary election majority;
- That Trump beating Cruz in all five March 15 states would neither be unexpected nor indicate a fundamental change in the dynamics of the race or its eventual outcome;
- That, nevertheless, such a victory, caused in substantial measure by Rubio’s decision to continue his campaign, might create the general perception that the race is over, depressing the enthusiasm of Cruz voters, funders, and volunteers and, thereby, turning the perception into reality; and, most important in the long run,
- That Republicans need to have a serious conversation about how much control over their party’s nominee they want to give non-Republicans. Leave even the proportion of open and closed votes the same and simply change their order, and the dynamics of the entire race would be different—and not for the first time.
As we wrote four years ago, the current primary system is well-designed to give non-conservative candidates an advantage over those who represent the party’s center and base. A number of factors contribute to this, but front-loading open primaries is one of the most important. Since 1988, when the present system debuted, it has produced a series of establishment over conservative primary election victories, but only one decisive victory in November, when George Bush won a “third term” for Ronald Reagan that first election cycle.
This time around, an anti-establishment, non-conservative candidate took advantage of this system. Meanwhile, Cruz has had to run the best conservative campaign in a generation to put himself in a position to stop Trump’s nomination. Nothing that happens Tuesday should keep conservatives from continuing to fight for him as if the republic depends on it—because it very well may.