Or “guessing” if you want to be more accurate.
The results of a new FDU-Public Mind poll hit email boxes this morning, Save Jerseyans, trumpeting a 14-point lead for Barack Obama over Mitt Romney in the Garden State (52% to 38%) among likely voters.
Farleigh Dickinson surveyed “706 likely voters statewide was conducted by telephone using both landlines and cell phones from September 6 through September 12.” Those 706 participants broke down by voter ID as follows:
Democrat/Lean Democrat: 47
Republican/Lean Republican: 37
But is FDU’s sample realistic? CNN conducted exit polling for the New Jersey U.S. Senate race in 2008; this is who they found voting at the polls based on a survey of 1,476 respondents:
I’m not accusing the Public Mind folks of knowingly conducting a bad poll, Save Jerseyans. This is simply another frustrating example of the primary challenge associated with accurate polling: predicting the appropriate turnout model! Few believe Obama isn’t winning New Jersey by a comfortable margin at the moment (defined by me as at least a 6-7 pt lead statewide). But is he up by only 7 points (that’s what the latest post-DNC Q-poll found) or 14 points as determined by FDU?
And for the record, the aforementioned Q-poll’s dissenting opinion was arrived at by surveying (1) far fewer Democrats and (2) many more independents:
A win is a win in the electoral college (in most states, anyway). That said, knowing how Obama is performing in one of the country’s more Democratic states can help illuminate the state of the race nationwide. The size of Obama’s margin will also affect races down the ballot.