Campaigns and elections

Likely Campaign Effects in 2008

Oct 6 '08

Political scientists routinely emphasize how the “fundamentals” — the economy, war, presidential approval, etc. — affect election outcomes. The implication is that every twist and turn of the campaign may not matter, or at least matter less than heavy-breathing cable news shows suggest. See Brendan Nyhan’s excellent discussion of this, as well as previous posts on this blog, e.g., here.

But none of this means campaigns never matter. And this graph from Sunday’s New York Times suggests when they do:


Obama is now vastly outspending McCain in Florida, and pulling ahead of him in Pennsylvania too (see here). This matters because disparities in the balance of communication can affect vote intentions. When Obama’s message drowns out McCain’s, he is likely to benefit.

There is an important precedent here: the 2000 election. In their analysis of this race, Richard Johnston, Michael Hagen, and Kathleen Hall Jamieson find that in the final 2 weeks of the campaigbn, Bush’s advertising dwarfed Gore’s in the battleground states:

bq. In the last week, a resident of a competitive state was more than half again as likely to see a pro-Bush ad as a pro-Gore one.

Did it matter? Yes, it did:

bq. Most eye-catching, however, is the very end, when the negative balance in fiercely contested DMAs [media markets] — all in or adjacent to states that would be Electoral College pivots — depressed Al Gore’s vote share possibly by four points among the least committed persons.

The fundamentals predict an Obama victory. But campaign dynamics may also facilitate this outcome, if Obama can consistently outspend McCain in these closing weeks of the campaign.