Thursday, August 6, 2009

What Will GOTV Look Like Ten Years From Now?

Part 9 of our 10-part series: 21st "Century Campaigning"

Get-Out-the-Vote (or GOTV) operations have changed dramatically in the last ten years, and – chances are – they’ll change even more in the next ten.

GOTV has always existed in some form or another throughout American electoral history. But modern GOTV methods really began to take shape across the pond. In the United Kingdom’s 1945 general election, Labour Party MP Ian Mikardo and his campaign developed a system in which the Reading constituency he represented was canvassed for Voter ID. The campaign then generated lists of supporters called “Reading sheets” for volunteers to remind to vote on Election Day.

The Reading System also began a GOTV tactic known in the U.S. as “poll watching” – in which campaign workers watch to make sure that voters identified as supporters have made it to the polls already and can be crossed off the list. If not, they get reminded to vote again.

In the United States, national GOTV operations really started to be noticed by the media following the 2000 general election. It was a year where the vote between Al Gore and George W. Bush was so close that getting supporters to the polls could make all the difference in many states.

That fact prompted the GOP to make important strides in the way GOTV was to be done in the future. They developed a new system they labeled “the 72-Hour Program” for the 2002 midterm elections.

From a 2003 Washington Post article on the new program’s implementation:

The 72-Hour Project was born of necessity after the 2000 election, when Republicans discovered that Democrats had done a better job of getting their voters to the polls in one of the tightest presidential races in history.

With prodding from White House senior adviser Karl C. Rove, White House political director Ken Mehlman and RNC Deputy Chairman Jack Oliver, the party undertook a top-to-bottom review of its get-out-the-vote operation, poured more than $1 million into more than 50 experiments to test how best to reach out to voters and then methodically set about implementing their findings in the midterm campaigns.

It paid off. In the subsequent 2002 and 2004 elections the GOP significantly drove up turnout among the Republican-friendly Evangelical community and scored some important victories for their party.

Still, the efficacy of such extensive GOTV programs is of considerable debate. Yale University has devoted an entire website to research done on the effectiveness of GOTV operations.

And Democratic strategist Mark Mellman argues that putting such weight on GOTV really only makes a difference when a race comes down to a couple thousand votes or so.

From a 2006 piece he wrote for The Hill:

“Experiments on turnout by [famed political scientists] Alan Gerber and Donald Green suggest that the most effective means of increasing turnout raise it by less than 10 percent — and that’s for people who get canvassed in person. None of this is to suggest that GOTV efforts are not valuable. When 2000 or 200 votes decide an election there is no question that GOTV efforts can make all the difference in the world. But again, that is simply not the case that is being argued by GOP operatives.

Can’t micro-targeting help them achieve spectacular successes? Anyone who has ever modeled data knows there is much more salesmanship than science in Republican claims about these efforts. Our firm and others on the Democratic side have been using these models for half a dozen years or more and we know they can make our efforts much more efficient; expand our GOTV and persuasion universes; and provide message guidance. So when races are otherwise marginal, the lift models provide can make all the difference between winning and losing. But no model is going to turn what would otherwise be a 5-point loss into a victory.”

With this idea in mind, how will campaigns adjust their GOTV strategy so it only counts if a race is going to be close? How will they plan such an important operation ahead of time if they don’t even know if it will be worth it?

And if these questions don’t make campaigns rethink how they do GOTV, maybe some recent trends will.

As we mentioned in our post on Monday, early voting might change a lot of campaign strategy in the next few elections. In fact, when polls showed then-Senator Barack Obama with a substantial edge among early voters, we commented on the importance behind the Obama campaign’s efforts.

“The real advantage of having supporters vote early is that the campaign can focus more time and energy on supporters that vote less frequently (because they are less politically engaged) in the final 72 hours.

So have early voting laws made October a month long GOTV period? It would certainly appear so, as this tactic is sure to be used in presidential campaigns to come.”

But if campaigns are going to devote time and energy to such a “Get-Out-the-Early-Vote” operation, what will that look like? Will they rely on more time-effective methods to convince folks who were already going to vote for them to do it early – methods such as email, robo-calls, text-messages, and direct mail? Or will they actually redirect volunteers from important Voter ID operations to make calls and canvass these supporters?

Ultimately, these aren’t questions that can be answered yet with any certainty – but they are questions that campaigns will have to answer for themselves in the next ten years.

The only thing that is certain is that GOTV will continue to evolve and is likely to look much different in 2020 from what it looked like in 2008.


Stephanie Maier said...

Back in 2002, I was fortunate to be the 72-hour Director for Ralph Reed and the Georgia Republican Party, where the success of the program significantly contributed to that year’s historical wins for Georgia.

Your blog asks, “With this idea in mind, how will campaigns adjust their GOTV strategy so it only counts if a race is going to be close? How will they plan such an important operation ahead of time if they don’t even know if it will be worth it?”

While the media covers elections in broad strokes – blue states, red states, and national percentages and straw polls – all elections are essentially local. Meaning, that although media coverage can impact public opinion, grassroots operations should be analyzed on a county by county basis, not on a state-wide or national basis. From that perspective, you CAN predict ahead of time if a race is likely to be close or not: counties tend to be primarily Democrat or Republican. Using public voter registration records, you can determine the location and the number of voters in your party’s/candidate’s base, the location and the number of voters in your opponent’s base, who has the larger base, the number of undecideds/no-party-affiliation registered voters, and even the voting history of each voter in terms of the number of elections they have voted in to better understand how likely they are to vote in this election. Also, by looking at the number of times each voter participated in a primary, you can determine the most likely way they voted in the general election. With all of this analysis at the county level, you can determine who has the larger base and which way the undecideds are trending in various counties, and thereby determine how many votes you need to get in each county and in the state overall to win that state. I don’t know if that’s clear in such a brief explanation, but my point is that there’s a numerical science to it that’s based upon actual voting data, and hard campaign strategists know how to study the numbers.

This is a perfect example of why a popular person such as Donald Trump or Ross Perot in the 1990’s was ill-advised to run as an Independent, because in these two examples, they can (and in Perot's case did) split the vote amongst conservative followers, effectively winning the election for the Democrat. This example can also be predicted in a county-by-county approach analysis.

The final point of all this is that most elections should be approached as though every vote counts, because close elections in certain counties can collectively impact the overall final results of the state and thus, the nation. Good strategists, therefore, don’t rely on the media to tell them whether or not their candidate is ahead or behind. This serves as a barometer, but the strategy can and should be much more exact than that.

Ben Donahower said...

Depending upon any given state's policies on voting by mail, and while far less likely perhaps even voting online, I also see GOTV being more distributed over time and adopting microtargeting principles as well.