There are polling results for both candidates to brag about today, but
overall the race seems to have entered a fairly stable phase.
In Florida, a new survey from Public Policy Polling
has John McCain with a 3-point lead. This is an improvement for McCain
from the June edition of this poll, when he had trailed by 2.
Meanwhile, in New Jersey, Rasmussen has Barack Obama ahead by 10 points, after having led by 3 points in July.
I lump these two polls together because in each case, looking at the immediate trendline is a little bit deceptive.
The preponderance of polling and demographic evidence -- including
previous surveys from these firms -- has suggested that John McCain
holds a small lead in Florida, and that Obama holds a larger,
relatively safe lead in New Jersey. So this may very well just be a
case of the numbers reverting to the mean, rather than any kind of
organic movement. Likewise, in New York,
Rasmussen shows Obama with a 19-point lead. This is down from a massive
31-point lead that he held last month -- but more consistent with other
polling of the state, including Rasmussen's numbers in April and May.
Still,
these numbers may impact electoral strategy at the margins. The McCain
people have been smart enough to avoid a serious investment in New
Jersey, which is a huge money trap. The Obama campaign, on the other hand, has been serious about trying to flip Florida,
which is also a notoriously expensive state to advertise in (lots of
old people = lots of people watching TV news). As I've elucidated
before, having gone this far in Florida, it would be premature for the
Obama campaign to pull out now -- by any measure it's a winnable state.
But we're still talking about a fairly narrow set of conditions under
which Florida matters. The basic flowchart for this election looks as
follows:
1. Can McCain win Michigan? If so, McCain is very likely to win the election.
2. If McCain loses Michigan, can Obama win Ohio? If so, Obama is very likely to win the election.
3.
If Ohio and Michigan are split, can Obama win Colorado or Virginia? If
so, Obama is very likely to win the election, having essentially to
pick off just one or two smaller states West of the Mississippi (Iowa,
New Mexico, Nevada, Montana) while perhaps also having to defend New
Hampshire.
These remain the paths of least resistance in this
election. It is not coincidence that Obama has two winning paths to
McCain's one, which matches the roughly 2:1 favorite that we make Obama
in this election.
States like Florida, Missouri and Indiana
represent break-glass-in-case-of-emergency states for Obama; McCain's
analogous states are probably Pennsylvania, Oregon and perhaps
Minnesota. If these states have become must-wins, we are talking about
a scenario where a candidate is facing third-and-long: not an
unconvertable position by any means, but also not the one they'd prefer
to be battling from.
--Nate Silver