The Efficient Obama Campaign?

by John Sides on January 8, 2013 · 1 comment

in Campaigns and elections

In response to my post, a Monkey Cage reader emails these three thoughts about the Obama campaign and its importance in 2012:


1) The difference really is just 0.2 percentage points, and that was all the edge that was possible with the diminishing returns of additional communication in such a crowded information environment, but it’s not the effect that will be statistically significant with 50 observations.


2) That the story is less about Obama than it is McCain and Romney, and McCain’s campaign effects were that much less differentiated between battleground and non-battleground states than were Romney’s. After all, it’s basically the same set of states in 2012 as 2008.  Some qualitative leverage on this idea might be gained by looking at the states that were battlegrounds in 2008 but not in 2012 (e.g. Indiana), as I don’t think any went the other way.


3) It’s not about relative vote share in a state, it’s about campaign efficiency. A great deal of what has come out about the Obama campaign focuses on optimization, such as their tool to reach their target audiences with less money. This is a point I made on twitter earlier when people were comparing dollars spent on both sides, including SuperPACs. Sure, part of that is campaigns getting lowest available rates and independent expenditures paying a fortune per point, but part of it is also about using information about where target voters are watching cable tv to get more efficient buys. I think that the spot market buys are generally not in the CMAG data you’ve used, but that’s something worth checking.  In such a world, the two sides may get to about the same place, but one side is going to spend less doing it, which I think is consistent with the latest spending totals, though I’ve not seen the all of the final numbers.

I thought the third point was particularly interesting.  After all, Moneyball as it was practiced in baseball was not just about using data but about being more efficient—getting better players for less money.


Points 1-2 may be right too.  Again, stay tuned.

{ 1 comment }

wheeler's cat January 9, 2013 at 12:07 pm

bzzzt wrong again. there just aren’t enough white voters left as a percentage to deliver a national election. if it was a snake it would have bit you Sides. Romney won a hefty 59% of the white vote and it wasn’t enough. 2012 was the tipping point for the demographic timer.
White vote share of the electorate was 72% in 2012, and projects to be ~70% in 2016.
The GOP will never win another national election as long as the GOP base is 88% non-hispanic caucasian.

Comments on this entry are closed.

Previous post:

Next post: