Prepare to be confused, especially if you read Associated Press reporter Kathleen Ronayne's recent contention that UNH pollster Andy Smith claims that Democrats enjoy a three percent advantage in "voter registration" in New Hampshire.
When I showed reporter Ronayne the data from the Secretary of State's office, data proving that just the opposite is true, she checked her notes and insisted she had quoted Smith correctly.
The data I showed her was that which I reported here two weeks ago, numbers showing the Republican edge was 3.2 points.
In fact, in numbers released by Secretary of State Bill Gardner late last week, the Republican edge has actually "grown" to 3.8 points.
These are actual numbers, mind you; no polling is involved.
New Hampshire Total Registered Voters--864,817
30.6% minus 26.8% equals 3.8% as the Republican advantage.
Just to be sure, I've double checked the percentages, but don't take my word for it; feel free to run the numbers yourself.
You'll find that far from having a 3% advantage, as reporter Ronayne claims Andy Smith as claiming, Republicans do in fact have a 3.8% advantage, 30.6% to 26.8%.
Let me run the numbers a third time.
They're not going to change.
The Republican advantage ticked up a bit since far more "undeclared" September primary voters opted for a Republican ballot, but that can't explain much of the difference because approximately 85 percent of the Undeclared voters reverted to their undeclared status prior to leaving the polls.
I suspect that what Smith was telling Ronayne was that when his pollsters ask voters to self-identify themselves, he finds a three percent Democratic edge (that's what Gallup finds--note this chart), but that's clearly not what the Secretary of State's actual numbers are.
Smith goes on to tell Ronayne (that's not an easy name to type correctly five times!) that Republican entusiasm should more than make up for the Democratic three point edge which in fact does NOT exist.
If Smith's polling data is based on a three point Democratic edge, rather than a 3.8 Republican edge as the actual numbers indicate, we should knock six to seven points off his totals for all Democrats in last week poll results. The would put Guinta up two or three points on Carol Shea Porter; Scott Btown would be virtually dead even with Jeanne Shaheen (as many other polls are showing); and Walt Havenstein would be within seven points of Governor Maggie Hassan (there would still be no hope for Marilinda Garcia whom Smith has down by 23 points).
For accurate polling results, any pollster will tell you, the starting point is to get the model right, and when a pollster is off seven percent with what he tells an Associated Press reporter, the old Dr. Henry Lee line from the O.J. Simpson trial comes to mind...