Was Arlen Specter or Blanche Lincoln the easier primary target?
Specter would seem to be the obvious answer. He had been a Democrat for barely a year; Lincoln has served her party in Congress since 1993. But, polls showed that Specter was actually thought more highly of by members of his own party. Rasmussen polling in April found Specter with a 64-33 favorability score among Democrats -- and in May, they found Specter with 67-31 favorability among likely voters in the Democratic primary (even though their poll showed Sestak ahead). By contrast, Rasmussen had shown Lincoln's favorability among Democrats at 55-37, in a poll also conducted in April. Public Policy Polling also had marginally worse numbers for Lincoln. A February poll found her approval at just 51 percent among Democrats, with 35 percent disapproving; Specter, in an April poll by the same firm, was a couple of notches better, at 53-30.
But Specter lost to Joe Sestak, and did so by a fairly solid 8-point margin. Lincoln survived, and actually improved her numbers somewhat between the May primary and Tuesday's runoff. Certainly, it is right to point out that this was hardly an unmitigated failure on the part of Bill Halter; a close loss should not be interepreted much differently than a narrow victory. Nevertheless, the 12-point gap in the performance of the challengers -- Sestak at a +8, and Halter at a -4 (based on last night's results) -- is not trivial and is worthy of consideration.
Some of the difference, clearly, is that Sestak ran a very strong campaign in the closing stages, whereas Specter stumbled his way to the finish line; Halter's campaign, on the other hand, sometimes had the feeling of a story with a good wind-up but lacking a punchline. But the composition of the Democratic primary electorates in both states was also very different.
Readers of major liberal blogs, who were active on behalf of both Sestak and Halter, are mostly white, college-educated, and liberal. Not all of the readers, by any means, but most -- that's the prevailing demographic for political and most any other type of blogs.
We can formulate a reasonable (although imperfect) estimation of how "bloggy" a Democratic primary electorate is by looking up its exit polls from 2008 and multiplying together the fraction of voters who were white, who were liberal, and who were college-educated. By this measure, Pennylvania ranks toward the upper half of the list, and Arkansas toward the bottom. Connecticut, where Ned LaMont was successful in his primary challenge against Joe Lieberman in 2006, also ranks highly by this metric.
But what about the unions -- which were much more active in Arkansas than they were in Pennsylvania (where, in fact, many of them endorsed Arlen Specter, although lukewarmly)? Unfortunately, they're not much help in Arkansas. In fact, Arkansas has the lowest fraction of union members from among the 50 states -- just 4.2 percent of the overall population, or 8 percent of the Democratic primary electorate, according to 2008 exit polling. In Pennsylvania, by contrast, 19 percent of Democratic primary voters in 2008 were members of unions.
The coalition of blog-friendly and union-friendly voters, in other words, had about as little influence as it possibly could in Arkansas; the degree of difficulty was much higher there. Even if Lincoln seemed like an easy target, to have pulled this off would have been really something.
I didn't follow much of the messaging that the Halter campaign used in Arkansas, and so I'm not in a position to criticize or commend it. Certainly, it is possible to use one set of messaging with your national activist base in order to raise cash and awareness for your campaign, and another with the voters back at home. But, relatively esoteric causes like EFCA or the public option were not liable to carry the day in Arkansas in the way that they might in Connecticut or Oregon. Even some red and purple states, particularly those in the Mountain West, which are fairly moderate overall, nevertheless have primary electorates with a lot of very liberal liberals.
If unions and the netroots want to find success in these more difficult environments, they will need to find coalition partners like African-American or Hispanic groups. And they'll need to keep in mind that the reasons they don't like someone like Blanche Lincoln may not coincide with critiques that local voters have.