Sunday, October 14, 2012


Does anyone know how Nate Silver calculates his "Democratic bang for the buck" index concerning where a $2500 contribution to Senate races will have the most impact?  There isn't an explanation anywhere on the site, as far as I can tell.  Which is unfortunate, because if it's calculated well, it could be very useful.  I suspect it isn't, though, because it's giving some seriously weird results.  At this point, it says the 4th best place to give money is Nebraska, where we're going to lose, and the 6th best place is Texas.  Nevada, Montana, and Wisconsin, all of which strike me as far better donation options, are in places 7-9.

Anyway, top of my list right now is probably Heidi Heitkamp in ND, as a recent Mason-Dixon poll has her in a 47-47 tie with Rick Berg.  But I'll wait a couple days to check out the October 15 FEC filings, as I'd like to see how much cash the campaigns have before I give.  


Daniel G. said...

He did a post about it at one point - I believe it factors in the competitiveness of the race as well as the cost and reach of the media market.

ikl said...

I share your intuitions on this. I love Nate Silver, but his Senate forecast has had a number of wierd quirks this year. For example, overrating the Dems chances in NE and underrating them in ND. I think that the problem is not enough public polling (and especially not enough quality public polling). So even if his bang for buck formula makes sense in theory, I'm not sure that the race projection numbers that he feeds into them are any better than your sense of the races.

low-tech cyclist said...

I'm not sure about Nate's Senate rankings either.

I agree with you that Heitkamp should be a lefty's first choice to give money to right now, among Senate candidates.

After that, my first-tier choices would be Jon Tester (MT), Shelley Berkley (NV), and Richard Carmona (AZ).

After that, Joe Donnelly (IN) and Chris Murphy (CT). I don't understand why Murphy hasn't managed to pull away in this one.