A cynic might say that tonight’s Republican contests in Minnesota, Colorado and Missouri deserve an asterisk. In Minnesota and Colorado, which will hold caucuses, voters will pick their preferred presidential candidate in a nonbinding straw poll, while picking delegates to county and regional conventions in a separate vote. In Missouri, no delegates are on the line at all; the state will hold a separate caucus for that purpose on March 17.
The results, nevertheless, will provide an important test of how robust Mitt Romney’s coalition is on less favorable terrain than in states like New Hampshire or Nevada. And they could potentially revitalize the campaign of one of Mr. Romney’s opponents, Rick Santorum.
Nor should one go too far in dismissing the results. The process that Minnesota and Colorado use, holding separate votes for presidential preference and delegate selection at their caucuses, is essentially the same one that was used in Iowa. Missouri is a more debatable case, but as the first primary of any kind held in the Midwest — perhaps Mr. Romney’s weakest region — it may tell us something about how states like Michigan and Ohio are likely to vote when they hold key primaries on Feb. 28 and March 6, respectively.
Polling and demographic evidence suggests that Mr. Santorum is the favorite in Missouri, while Mr. Romney is favored in Colorado. Minnesota could be the closest contest, but probably leans slightly toward Mr. Santorum.
The Polls
We haven’t run forecasts in any of the states. The FiveThirtyEight forecast model was “trained” on past cases in which at least three different pollsters were active in a state in the closing days of the election. We cheated slightly to run a forecast for Nevada, where two firms issued polls, but with just one pollster testing match-ups in each of tonight’s states, my view is that we’re better off looking at the polls and the other evidence in a more holistic way.
The polling firm active in each state is Public Policy Polling, which released a set of survey results on Tuesday morning. Their polls give Mr. Santorum a 13-point lead in Missouri and a 9-point lead in Minnesota, while Mr. Romney has a 10-point lead in Colorado.
Although Public Policy Polling has had good results in elections so far, leads of this magnitude — about 10 points — are not especially safe, especially with just one poll in the field, and especially in the caucus states. However, a demographic analysis of the states tends to give credibility to Public Policy Polling’s results.
Demographics
Harry J. Enten, writing for The Guardian, has found a strong correlation between the number of evangelical voters in the state and the share of Mr. Romney’s vote. That could be problematic for Mr. Romney in Missouri and Minnesota, where the share of evangelicals should be relatively high.
Correlation is not necessarily causation, and conclusions based on just five data points can sometimes prove to be specious. In this case, however, there is evidence that the religious orientation of voters has a direct rather than incidental relationship with the outcome of the vote. According to exit polling in the first five states, Mr. Romney has won an average of 30 percent of the vote among voters who describe themselves as born-again or evangelical Christians, versus 45 percent among those who do not. Conversely, Mr. Santorum has won 22 percent of the vote among evangelicals, versus just 9 percent among other voters.
Another strong relationship so far is that Mr. Romney has performed well among wealthier voters, while having mixed results among voters with lower incomes.
We can estimate the income status of each state by looking at the share of John McCain voters whose households made $50,000 or less in 2008, according to that year’s exit polls.
Missouri, by this measure, is the most working-class state to vote so far. In 2008, 39 percent of Mr. McCain’s voters made $50,000 or less, well above Mr. McCain’s national average of 32 percent. Colorado, conversely, is a wealthy state; just 23 percent of Mr. McCain’s voters made less than the $50,000 threshold in 2008.
Minnesota is something of a mixed bag. Although the state has above-average incomes overall, that is not necessarily true among Republican voters; about one-third of Mr. McCain’s voters made less than $50,000 in 2008, close to the national average.
Google Search Traffic
Another thing worth looking at is Google search traffic, which gives evidence about the level of interest in each candidate in each state. This would have told you something about the vote in past states. In Nevada, for instance, where Mr. Romney and Ron Paul performed more strongly than their national polls, their search traffic had been strong, whereas it had been relatively weak for Newt Gingrich and Mr. Santorum.
In the chart below, I’ve looked at the number of Google searches on each candidate’s full name in each state over the past week, and then compared it to a baseline, which is calculated by taking the median amount of search traffic for a candidate in the 10 largest states.
The candidate who does the best by this measure is Mr. Santorum. In each of the three states voting tonight, his name is generating about twice as much search traffic as it does in the baseline states. That suggests that voters are taking a serious look at Mr. Santorum in these states, and that the fact that he is well behind Mr. Romney in national polls is not of that much significance.
Mr. Gingrich does poorly by this measure; he’s getting essentially no more search traffic in Colorado, Minnesota and Missouri than he is in inactive states like California or North Carolina.
Mr. Romney and Mr. Paul are somewhere in between, getting above-average traffic in Colorado and Minnesota but not in Missouri, where Mr. Santorum’s numbers are much more impressive and where Mr. Gingrich is not on the ballot.
Even though we’re not officially issuing forecasts for these states, we can apply the same technique that we have in the past and consider stronger and weaker scenarios for each candidate.
Mitt Romney
Strong scenario: He wins all three states, although Missouri is very close.
Most likely scenario: He wins Colorado by a clear margin and loses Missouri by about 10 points; Minnesota is very close but he loses there once all votes are counted.
Weak scenario: He takes a very narrow loss in Colorado in addition to Minnesota and Missouri.
The equation here is pretty simple: the more states that Mr. Romney loses, the more evidence we will have that his position is vulnerable and that the race could go on for some time.
Colorado, however, is unlikely to be a loss for Mr. Romney. Although his 10-point lead in the Public Policy Polling survey is not all that strong, the demographics look reasonably good for him, especially since perhaps about 10 percent of the voters there will be Mormon, making up for deficiencies among other groups. The flip-side is that if Mr. Romney were to lose Colorado, it would suggest deep problems for his campaign.
At the other end of the spectrum, if Mr. Romney were to sweep all three states, a strong case could be made that he has the nomination essentially wrapped up. So he has a fair amount on the line tonight.
Rick Santorum
Strong scenario: He wins all three states, although Colorado is very close.
Most likely scenario: He wins Missouri by a clear margin and Minnesota narrowly and loses Colorado by a clear margin.
Weak scenario: He gets shut out.
If Mr. Romney were to endure some losses tonight, he would probably prefer that these came at the expense of two or more candidates, leaving the rest of the field somewhat disorganized. Mr. Santorum, however, appears to be the favorite in both Missouri and Minnesota and probably has the best upset chances in Colorado.
Mr. Santorum has drawn closer to Mr. Gingrich in national polls, and a strong night tonight might be enough to propel him into second place, even though the contests are receiving tepid interest from the news media.
If Mr. Santorum performs well, it will be important to see whether influential Republicans get behind his candidacy or instead start to attack him. On the one hand, Mr. Santorum is almost certainly preferred by Republican elites to Mr. Gingrich. On the other hand, he is probably a more serious threat to Mr. Romney than Mr. Gingrich was, so if Republicans are determined to have Mr. Romney as their nominee, their praise of Mr. Santorum could be faint.
Newt Gingrich
Strong scenario: He gets a surprise win in Minnesota.
Most likely scenario: He finishes in third in both caucus states. (Mr. Gingrich is not on the ballot in Missouri.)
Weak scenario: He finishes dead last in either Minnesota or Colorado, while Mr. Santorum has a strong evening.
Unless he pulls off a big upset in Minnesota, it is hard to see what good comes for Mr. Gingrich out of this evening. If Mr. Romney has a strong night, he will be well on his way to the nomination. If Mr. Santorum does instead, he has a good chance of emerging as Mr. Romney’s major rival. Only a split decision — and one in which Mr. Gingrich beats expectations by overperforming his polling — would leave his condition improved going forward. An especially weak showing, after disappointing results in Florida and Nevada, would have his campaign on the ropes.
Ron Paul
Strong scenario: He gets a surprise win in Minnesota and finishes ahead of Mr. Gingrich in Colorado.
Most likely scenario: He is not too far out of the running in Minnesota but is well behind in Colorado and Missouri.
Weak scenario: He finishes out of the running and worse than his polling in both Colorado and Minnesota, calling into question whether he has an advantage in the caucus states at all.
Nevada was an odd state for Mr. Paul. He beat his polling there and got a respectable 19 percent of the vote. But the caucus states are supposed to be a strength of his, and he will need to do better than that to secure enough delegates to have a real influence upon the Republican race.
Mr. Paul’s supporters sometimes blame the news media for its lack of attention to their candidate — and they make a reasonable case, in my view. But that case will be weakened if he cannot perform well in states like Colorado or Minnesota, where none of the candidates have received much news media attention, few ads have been running and a candidate with enthusiastic supporters and a strong ground game should have the potential to do well.