No, the party id survey came in at R+2.6... But he uses a D+3 sample.
Here... http://www.rasmussenreports.com/public_content/archive/mood_of_america_archive/partisan_trends/summary_of_party_affiliation
OK, thanks for clarifying.