To poll or not to poll?

There is some debate about whether or not a poll would be good for Scott Brown's Senate campaign. One of Jim Geraghty's readers summarizes the argument against the poll nicely:

The last thing we really need here is a poll on the race. There hasn't been any; a poll showing Brown very close would wake up the Democrat machine here. A poll showing him getting crushed could depress his momentum. Right now I'd give Coakley the edge on registration alone, but I don't see her winning by more than 10 percent.

William A. Jacobson adds some further comments along this line:

Any poll now would rest mostly on name recognition, and Coakley wins the name recognition contest today. So any poll taken now will show Coakley with a substantial lead which may not reflect the vote on January 19. Brown's momentum is relatively new, and there are bound to be further gains in the coming two weeks.

Polling also would be particularly difficult for this special election, where turnout could mean everything. A lot of people who say they will vote may not do so, particularly for a candidate with tepid but wide support.

Those are all valid points.

Part of the disagreement about polling derives, I think, from a disagreement about the strategy Brown should use to win in Massachusetts. There's a case to be made for an under-the-radar campaign, in which Brown slowly chips away at Coakley's lead, while she (mistakenly) assumes she has this race in the bag. With the Democratic machine unmobilized and without any national attention, Brown would be able, so the thinking goes, to come out on top on election day. The political world reels in shock, Republicans across the country break out in wild parties, and so forth. That's a possible, and somewhat credible, scenario.

But there's another scenario, too. Under this other one, there are so many institutional factors against Brown (power players in Massachusetts, Democratic registration advantage, etc.) that he would need to run a transformational campaign against them. An under-the-radar campaign would be electoral death, as Brown is crushed beneath the institutional inertia of Massachusetts politics.

I think there are elements of truth to both strategies, and, at the moment, I tend to believe that Brown needs to appear competitive in order to have a serious chance at winning. A public poll showing him closing on Coakley might do that. An infusion of donations could also do that. Perhaps one of the best things for Brown's campaign could be a private poll, one showing him with a chance, that could then be passed around to the RNC and various donors. Special elections have a very weird dynamic. There's almost no way Brown could win an under-the-radar campaign if this were an election on a November of an even-numbered year; maybe there's a chance he could win that way this time.

But Brown will need more than Republicans to win this one; he needs independents and needs to find a way to bring them to the polls (and they're not going to show up for what they regard as a lost cause). Look at these numbers from Jim Geraghty:

But to illustrate how tough the odds are for Brown, let's pretend that every registered Republican in the state, as of 2008, shows up and votes for him. And let us pretend that the independents split evenly, and that only one third of the state's Democrats show up and vote for Coakley.

Under that scenario, Coakley still wins by about 1,045 votes. That's how steep an uphill battle Brown faces in this race.

Brown's biggest hope is to get a number better than 50% from the independents. Independents constitute slightly over half the voters of Massachusetts, and, if Brown can get a good majority from independents, a high turnout from them and the GOP could push him over the top. (Sean Trende plays out the numbers a little more here.) Partisan intensity alone won't get Brown to Washington.