So for the last few days I have been making calls for the GOTV effort for Romney. In this process I have learned a few interesting things about “polling” conducted by the campaigns. I thought I’d share this with you.
We all know that both campaigns run push polls to attempt to define their opponent’s and their candidate’s image for voters. This discussion doesn’t include those polls, many of which are conducted by outside interest groups. Campaign polling early in year, or after the primary in a state, are conducted for two purposes. First like any other poll they attempt to determine which states are winnable and which issues appeal to specific voters. Second they gather data to allow micro targeting voters. All information collected is stored in a database which is fully personalized. In other words, if you have been called by either or both campaigns, that campaign knows exactly how YOU answered that poll. Voters get analyzed in 2 dimensions, certainly to vote, and solid pro, weak pro, undecided, weakly opposed, strongly opposed. In addition to how you answer questions about your likelihood to vote, public records are analyzed to determine how many presidential and non-presidential elections you have voted in, as well as (where available) in which primaries you have voted.
By September the campaign has determined, with about 90%-95% accuracy, if you are a “safe vote” or a “leaner”, or opposed and if you are a certain voter or a “low turnout voter”. BTW: If you have not voted in at least 4 consecutive federal elections you are automatically considered a low turnout voter. Based on data gathered over the last 8-10 election both sides know that, on average, it requires eight contacts (phone or door knock) to raise the likelihood of a low turnout voter showing up by 23%. At this point campaigns conduct two very different voter surveys. One is exactly what you would expect, general polling to see their margin of victory or loss among all voters. The other “survey” is solely of their own voters and leaners and solely to drive the GOTV effort.
Remember back in the late spring and early summer when the Obama campaign spent 15 million dollars “polling” the swing states? Some of that was push polling as part of the effort to make Romney unelectable, but a lot perhaps more than half was to build and update their database. The Romney campaign accomplished a lot of this prior to and during the primaries and then refined it prior to the convention.
So in September and October political calls and door knocks are targeted at low turnout voters who are safe and persuadable to contact them eight times. The other goal is to refine the database to prune any opposing voters, and former persuadable voters whom you’ve lost. You are also trying to eliminate voters who would vote for you but are unavailable, overseas military, or recently moved for example.
Two weeks before the election the process shifts to a final effort to improve your targeting. The last seven days is all about getting those low turn out voters to turn out and identifying your low turnout voters on your general voter lists for your poll watchers. On election day, you spend the morning with general turnout calling, and about 11:00 you start calling “morning” voters who haven’t shown up. The rest of the day is calling and “door knocking” people who your poll watchers haven’t seen yet to get them to the polls.
On the Republican side the GOTV effort this year, in terms of numbers, intensity, and organization, dwarfs anything they have ever done. Should this be a turnout election, I can’t tell you if it will be enough to win, but I can tell you that the parties, the establishment, and the grass roots have put any differences aside and worked tirelessly leaving nothing on the table. I hope it’s enough.
UPDATE by PoliPundit: Thanks for volunteering, UncleFred! You too can do what Fred’s doing. Sign up here.