Peter from Pasadena writes with an excellent question:
“I am hoping that at some point you might comment at the site on the election and polling impact of the micro-targeting and 72-hour turnout techniques that we have heard so much about the Republican Party using. These techniques have evidently been responsible for anomalous and ahistoric levels of GOP turnout when specifically and thoroughly applied in the last few election cycles. Are they powerful enough to drive surprise elections results on November 7? If so, to what degree?”
This is a really fantastic question. Unfortunately, it admits of little more than an anecdotal/intuitive answer. The effect of voter mobilization upon final vote turnout is something that has been under-studied in scholarly circles. I think this has to do with a lack of data. Offering a rigorous test of voter mobilization — one that makes a serious attempt to identify whether the apparent link between turnout and mobilization is not simply a product of spurious causality — would be difficult to do because parties, candidates, and outside interest groups do not offer the details of their programs to social scientists.
To appreciate this, consider what we would have to do to really test the effectiveness of mobilization. You would build a model that predicts the final vote in a district that depends upon a whole host of factors like demography, candidate spending, voter interest, etc. To inquire whether the 72 Hour Program makes a difference, you would include a measure of it in your model. Ultimately, your goal would be some kind of equation that predicts how a party’s share of the vote. For instance:
Republican Share of Vote in District = Baseline + Influence of Demographic Features + Influence of Candidate Spending + Influence of Voter Interest + Influence of Resources Dedicated to GOTV Effort
The idea here is that each of the factors on the right-hand side of the equation has an independent effect on vote choice. We would expect the GOP get-out-the-vote efforts to be positively related to GOP share of vote, holding all of the other variables constant, and that its positive relationship is not explicable by simply random factors (i.e. the difference between the reported effect and 0 is outside the “margin of error”).
The problem is that we cannot really “operationalize” the final variable, resources dedicated to GOP GOTV effort, because we lack the data. This is a general problem with the parties: they are public-private organizations, and only required to release financial data at a level of specificity that is much lower than what we need to “run” this model.
What is more, they are not required to release information on how they organize their operations so we cannot even necessarily use a measure like “RNC Spending In State” because who knows to what extent the state/local parties are picking up the tab for GOTV efforts. There is plenty of legal “money laundering” that goes on between party organizations.
So, we cannot say conclusively that the GOP’s 2004 efforts had a decisive effect, nor can we make a conclusive argument for 2006. We simply lack the data.
My intuition is that mobilization will make a difference, though not as much as it did in 2004 and 2002. Observing political actors gives us some prima facie evidence on this front. I tend to heavily discount the “conventional wisdom” of journalists/pundits because the nature of their jobs is to just offer endless pontification — day in, day out.
There is no consequence if a pundit is wrong. No real reward if a pundit is right. So, they can go down any randomly incorrect causal path and it will not make one whit of a difference in the world. Their job is just to “blah blah blah” all day, right or wrong.
Political operatives are very different. Unlike the pundit class, where there really are no stakes whatsoever, the stakes are high among politicians and their strategists. And I have noticed that all political operatives seem to be in awe of the GOP’s current program. The GOP views it as their secret weapon. The Democrats view it as that which could doom them.
Apparently, its force is so great that it induced Democratic National Committee Chairman Howard Dean, and Senate and House campaign chiefs Charles Schumer and Rahm Emanuel — three incredibly assertive and self-confident “alpha” males, who between them strongly hold two radically different visions of the future of their party — to reach an armistice. That is something.
More generally — voter mobilization is a long-standing tradition of American politics. If it did not work, I suspect that strategic politicians would have moved away from it long ago. So, the fact that we cannot demonstrate its efficacy via a statistical model does not mean that it is ineffective. Our inability is a testament to our lack of data.
As I mentioned, I think the difference will be less than it was in 2004 and 2002. The presumably dispirited state of those the program seeks to mobilize seems to me to necessarily reduce its efficiency (i.e. it will cost the GOP more money to be as effective as it was in the past).
Mobilization reduces the costs (and increases the benefits) of voting by reminding people to vote, by helping them get to the polls, by making them feel like they are performing a civic duty, etc. Accordingly, its effectiveness is predicated upon the voter’s assessment of the costs/benefits of the voting act.
If GOP voters are seeing lower benefits to voting because of dispiritedness, then the same amount of mobilization activity will be less effective, as the average voter needs “more” to get him to the polls.
Leave a Reply