Democratic Sen. Mark Udall (D-Colo.), left, and his opponent U.S. Rep. Cory Gardner (R-Colo.) watch a coin flip by moderator Brandon Rittiman to see who gets first question during their final pre-election televised debate. (Brennan Linsley/AP)

Sen. Mark Udall (D-Colo.) is the latest of a very, very, very long line of politicians to play the "we'll show 'em" card. We'll show those pundits, he declares of his campaign, which trails narrowly in poll after poll. The ground game, our turnout operation will make the difference! Such stories almost always end the same way. The candidate's confidence, real or affected, turns out to have exceeded the boundaries of reality. The turnout didn't happen. He lost.

But can it make a difference? Is there a possibility that Udall can make up a four-point difference by getting more people to the polls? And can Democrats who are worried about similarly small deficits in a series of tight Senate races bank on the much-hyped Bannock Street Project being their ace in the hole?

To answer that, we turned to Columbia University Prof. Donald P. Green, who, with his long-time collaborator Alan Gerber has studied and quite literally written books on turnout. Some of their research is collected at a lamentably out-of-date page on the Yale University system, but we wanted to ask Green for his thoughts about turnout-as-savior directly. Below is a transcript of that conversation, edited for clarity and length.


POST: One of the main questions that exists about these midterms -- and more broadly about elections in general -- is the extent to which voter turnout shifts can affect outcomes. If you had to come up with a rule of thumb, how much of a difference do you think strong, smart field can make?

GREEN: Sometimes the voter turnout part of the campaign is called the "field goal unit," perhaps derisively. It's not an altogether unfair characterization when you think about the following consideration. A very, very strong get-out-the-vote effect will be something in the order of 8 to 12 percentage points. That would be a hugely successful face-to-face canvassing operation.

But even if that operation is pretty darn extensive, say in the midterm elections, you might contact a third of the target voter population, and move them 9 percent of the way. That's an advantage of three percentage points. And that's not even three percentage points of margin, because it's just bulking up the people they consider to be their supporters by three percentage points. That's why to say two to three percentage points isn't a crazy thing, but it's a matter of guesswork.

Sometimes I think that voter mobilization campaigns flatter themselves to think that they can have that effect. Why? Because they're not doing a lot of the things that are considered best practices. They're making contact with things like direct mail, which is a stretch.

POST: It's been interesting to watch the emphasis that's been placed on GOTV and field since 2008. To what extent do you think campaigns at the national level really are being smarter about how they target their voters, and to what extent do you think it's spin?

GREEN: I think that they are getting better. When I first started doing this in 1998, campaigns had very little sense of what you would go about doing in the world of voter mobilization.

Nowadays, people have a sense of what's really going to work. It has to be authentic, genuine, as personal as possible. If it involves mass media, it has to at least implore people to vote; it can't simply be the crush of campaign activity. That's been pretty well dispelled, as has the notion that people don't vote because they don't know where to vote or what time to vote or what day to vote. Those are largely debunked notions.

Presidential campaigns have gradually moved from persuasion-oriented to mobilization-oriented strategy. They've done that in part because they're striking a deal in effect with their activists to be able to run a more ideologically motivated campaign, get more of the activists fired up, see more ground activity -- and pay less and less attention to the persuadable voters that might be in the middle.

POST: Let's say you're Sen. Mark Pryor (D-Ark.) today. You are probably trailing, at best close in a race. And let's say you're starting from scratch. What would you do for the last two weeks to maximize turnout in your base?

GREEN: If I were working in places where Democrats needed core voters to come out and vote, I would probably return to places that were quite successfully mobilized in 2012 and are overwhelmingly Democratic in elections -- heavily African American elections for example -- and try to do it one more time. You're not taking a person who has no interest in politics and is not registered. These are people who think of themselves as voters. They might be demoralized or disaffected, but it's not that they don't participate in elections. They just need to be pushed over the behavioral threshold to vote in this election.

In densely populated areas, canvassing would be your best bet. The problem increasingly is that phone campaigns are difficult, especially with people under 40. Even under optimal circumstances, they're not necessarily great. Very often, they're paid callers from a commercial vote bank, and they crank through those calls 50 or 60 an hour. It can be uninspiring. The more pizzazz you put it in it, the better.

There are other things that Democrats and Republicans can do, but especially Democrats. You can bring out the vote in areas that have early in-person voting. There you have the opportunity to bring people together on weekends. It's also the case in places where you have widespread or entirely vote-by-mail systems. You can have gatherings that are essentially voting activities.

POST: Have you seen any recent experiments around field that seem interesting?

GREEN: What I've seen a lot of is confirmation of long-standing propositions.

The tricky thing is to try to harness the new technology, the social media where things go viral in a hurry, to do something with that. In previous years, especially in the 2008 campaign that the Obama folks ran, they were genius about extracting contact information from social media websites. But that's a done deal. The next challenge is to have messages through those connections that actually mean something in terms of vote production.

Advertisements are ineffective, or have proved to be ineffective. The next frontier for both sides of the aisle is to try and figure out how to make that work.

POST: What's the best turnout operation you've ever seen?

GREEN: It's a tie between Bush-Cheney '04 and Obama for America 2008.

The Obama campaign had enormously high morale, and you wouldn't have got that from the Bush-Cheney campaign in 2004. But they were so ideologically fired up from the prospect of John Kerry as an opponent, they worked like crazy to produce votes. Their turnout operation in Ohio will always be kind of legendary. But the fact that the Obama campaign operation in 2008 was able to win in places like North Carolina and Indiana? I don't think anybody would have ever thought that they could do that.

It's hard to know what exactly is due to turnout, partially because both of those campaigns have remained so tight-lipped about what exactly they did. It would not be impossible to think that they produced very large numbers of votes, because the volunteer base was doing high-quality voter contact at a grand scale for weeks and weeks and weeks.

POST: Will the effects of field always be intangible? Will field always be something of a mystery, a black box -- or not?

GREEN: I lean towards the "it's not."

There have been something in the order of 45 to 50 canvassing experiments in the United States and other countries, left, right and center. They all produce results that are more-or-less in the same ballpark. If they're doing it with a sensible script, they're in the expected range.

That leads me to think that, as long as you're at the door and saying something appealing and sensible, voters take away a lot of the non-verbal communication. 'Oh, there's a guy at the door telling me that the election is important. He's sort of like me. Maybe the election is important.' I think from that standpoint, it has less of an intangible feel.