Secret Sauce at work in the North Carolina Senate race

One race: North Carolina Senate general between Republican Richard Burr and prospective Democrats Cal Cunningham and Elaine Marshall. Two polls: Public Policy Polling and Rasmussen Reports. Two markedly different results: Rasmussen shows Burr nearly 10 points higher than PPP does.

What’s going on?

PPP has Burr ahead of Cunningham 43-35 (MoE 3.6) and ahead of Marshall 43-37, Rasmussen has Cunningham losing 53-31 (MoE 4.5) and Marshall down (50-32).

PPP has Burr at about an 83% average lead probability and under the 50% danger line, while Rasmussen has Burr at a 98% average lead probability and at or above 50. However both have the pictured matchup of Burr and Marshall looking better for the Democrats. So the relative pairings agree, it’s just the absolute support of Democrats and Republicans that varies between the polls.

The difference appears to be that PPP polled “voters” (presumably 2008 voters) while Rasmussen sticks with his model of “likely voters,” which can vary from 2008. Rasmussen has been up front in saying he expects Republicans to have an advantage in 2010, while PPP’s voted-for-Obama crosstab better matches the 2008 results in which Democrats had better motivation.

Both methods are legitimate, and the fact that both have Marshall running ahead of Cunningham is consistent with both being honest. What both polls cannot be is correct. One of the two is making the wrong assumptions, leading to wrong results. If the disparity continues, come November North Carolina will be a great test of the two likely voter models. One secret sauce will win and one will lose.