People from the Analyst Institute, AFL-CIO, and Rock the Vote present some of their evaluations of various campaign initiatives in the 2008 cycle, and stress the importance of actual, data-based testing of various theories of what work and what doesn’t.
We are reminded of some old findings: robocalls don’t work, rushed calls don’t work, email doesn’t work for GOTV.
The general lessons are: the more personal, the better, always emphasize local affiliation and personal affinity when making voter contacts.
Some bits from the discussion:
People talk about ‘spill-over’ effects. Apparently, most forms of GOTV voter contact has a spill-over effect of about 60% on other members of a given household. So a campaign may not want to target one supporter living with four people who support the opponent.
There is some discussion of a mail piece that was used in Michigan, where targeted voters got a piece of mail before the election listing their own voting history and the voting history of 12 neighbors, a mention that voting histories are public information, and a line about a similar and updated letter being send out after the election. Obviously, the point is to shame people into voting. The results are mixed–the letter had a lot of impact, but also generated a backlash in the press, via mail and phone complaints, and the like.
A more general, and perhaps more interesting discussion was of under what conditions the findings of these experiments hold up. People express their skepticism on various counts, does over-saturation make a difference, how quickly does the effect wear off, and the general response from the presenters was two-fold, (a) there are always numerous conditions, “we could discuss those all day”, and (b) there are a million things more that needs to be tested (effect of race on canvassing, whatnot).
Regina Schwartz, who was on the panel from the Analyst Institute, pointed out at the end that it is in the best interest of anyone who works in, say, field, to get their results tested and evaluated, so that they are better equipped for internal budgetary fights.
rasmuskleisnielsen.net 