Rasmussen shows that Huckabee now “leads” Romney 28-25 in Iowa.  Like Obama, his “lead” is still within the margin of error, but as the latest symbol of his tremendous surge of support and Romney’s collapse it is significant (Rasmussen calls it a “stunning change”).  In Rasmussen polling, Huckabee has jumped 12 points during the month of November.  Before too long, pundits who have just finished writing, “Did Romney peak too soon?” analyses may start writing the same thing about Huckabee.  Now it’s time for fun with crosstabs!

There has been a lot of speculation about how Obama’s stronger support among first-time caucus-goers and younger voters, particularly college students, would affect turnout for him on Jan. 3.  (Some have noted that the Christmas holiday break actually works to Obama’s advantage because it spreads out the college students to their hometowns and boosts his representation in each part of the state.)  The assumption has been that younger voters and first-time caucus-goers, who are often the same people, are more unreliable and cannot be expected to show up in sufficient numbers on caucus night.  Romney has a similar problem.  For some inexplicable reason, young voters embrace Romney and prefer him over other candidates by a huge margin (he gets 45% among 18-29 year olds, compared to Giuliani’s 20 and Huckabee’s 15), but in every other age group, except 65+, Romney trails Huckabee by a statistically significant margin.  Huckabee leads among former caucus-goers 30-23, but trails among first-time caucus-goers 29-26; if turnout is going to be as anemic as expected this cycle, Romney may be in more trouble than it appears.  In short, if the students and first-time attendees don’t turn out for Romney, it is much more unlikely that he can win.  It is probably the case that Romney’s support is so high among younger voters because he has saturated their media market and his name recognition is much higher than many of the other candidates, which means that his broad-but-shallow support may be even more shallow than we thought. 

Huckabee also leads among most income groups , and Romney, strangely enough, polls best among <$20K earners (36%).  The only income groups Romney wins are the <$20K and $40-60K earners.  The more "downscale" the voters, the more competitive Romney is with Huckabee, which seems counterintuitive.  Among those earning $60K or more, Huckabee leads Romney by no less than six points.  Huckabee's populism may scare away the donors, but it doesn't seem to trouble the higher earners in Iowa all that much.  (Giuliani receives by far his strongest support among the >$100K earners at 22%, as does Paul at 9%, and so they have more of an effect on this group of voters, which could conceivably have opted for Romney if Giuliani weren’t in the race.)  Huckabee also does respectably well as a second choice at 16%, roughly even with Thompson and Giuliani and just behind Romney (21%). 

Where the Giuliani and Thompson voters (the next two largest blocs) go if either group is unable to reach the minimum level of support in any given district will probably determine the final outcome.  The shared interest of Giuliani and Huckabee in defeating Romney is well-known by now, so an unholy alliance between those two campaigns could be enough to propel Huckabee to victory.  Thompson can help Romney, but at 11% he doesn’t have enough raw numbers to put Romney over the top.  Besides, like Giuliani, his Iowa organisation is woefully weak.  The strength of his organisation may be what saves Romney in the end, if it can bring in enough of the disorganised Thompson and Giuliani voters.  Given Huckabee’s public, slightly harsh sparring with Thompson, it is unlikely that he will be the second choice of many of the latter’s supporters.