Polls Get Tighter, But Internals Show Terrain Getting Tougher For Deeds

First we had the Clarus poll released the other day showing that among Registered Voters Bob McDonnell was leading Creigh “What’s In Your Wallet?” Deeds by 5 percent.  Today, the latest Rasmussen poll out shows McDonnell’s lead among Likely Voters (including “leaners” — undecideds pushed to say who they’re currently leaning towards) is at 2 percent.  Undoubtedly, this race was bound to tighten up eventually as it appears to be doing (although given Deeds’ simply pathetic performance in today’s debate before the Fairfax County Chamber of Commerce, it may not remain close for long.)

While most reporters simply look at the headline of a poll press release and perhaps scan a few lines to get a sense of the horse race, not very many look at the internals of the polls and even fewer actually understand their significance.  There is much to criticize with the Clarus poll (polling Thu-Fri-Sat-Sun-Mon is generally not an acceptable sample by most standards – one can do Sat. polling but only during the day and you must still keep a close eye on things – and respectable pollsters NEVER poll on Fridays – NEVER) and even this latest Rasmussen poll (only a one-day sample).  But we here at Virtucon prefer to err on the side of caution and are willing to give these results the benefit of the doubt – for now.

All that being said, the internals for each of these polls show that Deeds has an extremely difficult road ahead of him and it is only getting tougher.  We now turn to Virginia Virtucon’s resident polling expert, Pedro’s Dwarf, for more as he reviews the internals from the Clarus poll…

I’ll tell you this, though.  If Deeds wins NoVA by only 5 points, he gets killed.  Here’s the 2006 exit polling:

GEO STRATUM CODE

Category	% Total	Webb	Allen
North	27	60	40
Southeast	19	53	47
East / Richmond	36	45	55
West / Shenandoah	17	43	57

Webb won NoVA by TWENTY points, en route to a statewide victory of 39 hundredths of a point.  (And yeah, I decided not to take a ton of time and do actual vote by regions manually, which would be more accurate.)  Leaving the other regions alone, and giving Webb a 53-47 lead in NoVA, Allen would have won by 51-48 (the Green candidate got a point statewide).

Here’s the 2008 exit polling:

Vote by Region

Total	Obama	McCain	Other
North (26%)	64%	35%	1%
Southeast (20%)	56%	43%	1%
East/Richmond(37%)	48%	50%	2%
Shenandoah-Southwest (17%)	40%	59%	1%

So Obama won NoVA by THIRTY-NINE points, en route to a six-point victory.

PD has some great points with that analysis and I have to agree that if Deeds can only manage a 5-point lead in NOVA among Registered Voters, he is going to get absolutely clobbered on Election Day when the statewide returns all come in.

Now, for some of the Rasmussen internal details:

Without leaners the numbers are McDonnell 47 Deeds 42. With leaners, 48-46.

GENDER –
Without leaners:
McDonnell gets 47% of men, Deeds 45.
Among women, McD 48, Deeds 40

With leaners:
McD 48 and Deeds 47 with men
McD 48 and Deeds 45 with women

AGE GROUPS –
Without leaners, Deeds captures the 18-29 age group and the 50-64.
McDonnell gets the 30-39, 40-49, 65+

With leaners, Deeds gets the 30-39, the rest remain the same.

AFRICAN AMERICAN VOTE –
Without leaners, McDonnell get 9% of the black vote and Deeds 77.

With leaners, McDonnell stays at nine, Deeds leaps to 88.

So, while these two polls show a tightening of the race on the surface, both of their internal numbers show a picture of a race that is becoming increasingly difficult for Creigh Deeds to be able to successfully navigate come Election Day.  In fact, Deeds will be lucky if he can manage to avoid a double-digit defeat at this rate.