How New Hampshire Voted

One of my favorite cross-tabulations in the exit poll is its breakdown of when voters made their final decision. In Iowa, Trump faded substantially among late deciders, leading to his second-place finish. Presumably he did better in New Hampshire? Well, sort of. In the table below, which is derived from the exit poll, I’ve grouped voters who decided “just today” or within the last few days into the “late decider” group and those who made a decision “in the last month” or “before that” into the “early decider” group. The results suggest that Trump didn’t do particularly well among late deciders, winning 22 percent of their votes. But he had a lot of voters who were loyal to him from the start:

CANDIDATE	EARLY DECIDERS	LATE DECIDERS
Trump	48	22
Kasich	10	22
Cruz	12	12
Bush	9	12
Rubio	9	12
Christie	4	10

What about Rubio? Presumably Saturday night’s debate cost him a lot of support? Yes, probably, although overall Rubio actually did slightly better among late deciders (12 percent) than early ones (9 percent). What may have happened is that voters who initially were intrigued by Rubio after Iowa backed away from him after the debate; the 11 percent or so of the vote he’ll get tonight is close to where he was in pre-Iowa polls of New Hampshire.

Fiorina and Carson are projected to finish a distant seventh and eighth, respectively, in New Hampshire. Despite their poor finish today, they have the support between them of about one in nine Republican voters nationally: In our latest national polling average, Fiorina had 2.5 percent and Carson 8.3 percent. So what happens if they drop out of the race? Cruz probably would benefit more than Trump would. The online pollsters at Morning Consult added up results from January polls it conducted among 5,456 Republicans and Republican-leaning independents nationally, asking for their first and second choices among the candidates. So far Morning Consult has published second-choice data for supporters of candidates who have already dropped out. They shared with us the data for Carson and Fiorina. Among Carson supporters, 24 percent had Cruz as their second choice, 19 percent named Trump and 10 percent named Rubio. Fiorina had far fewer supporters, but they might be higher leverage: 23 percent said they supported Rubio, 14 percent named Cruz and 5 percent named Trump. (Another 18 percent named Carson, and in this scenario those supporters would need to go to their third choice, or maybe skip voting.) Of course, these polls preceded the votes in Iowa and New Hampshire, and voters’ second choices could be even more volatile than their first choices are.

It’s not clear yet whether Christie will end his campaign. As I wrote earlier, Christie has been perhaps the most consistent Republican in the debates. He’s also a strong retail campaigner, which was evident here in New Hampshire. But Christie’s liabilities, from Bridgegate to his periodic deviations from conservative orthodoxy, were formidable also. He had plenty of opportunity to break out in New Hampshire and instead faded toward the end. It just doesn’t look like voters are buying what Christie is selling.