Democracy is a giant, on-going pain in the ass.

It is largely a pain in the ass because so few people really care that much about it. Nearly every American overwhelmingly supports democracy in concept; also, giving money to charity in concept; also, free speech for the Klan in concept. But in practice, democracy is tedious. It is tedious to run for office, it is tedious to hear from people running for office, it is tedious to vote for the person running for office. The goal of a political campaign is to make that second part as unavoidable and the third point as tedium-free as possible.

Depending. Sometimes campaigns kind of want voting to be as tedious and difficult as possible. This article isn't about that, about voter suppression and racism and America's long, slow, grudging decision that everyone should be allowed to vote and attempts to push that heavy train backwards. This is about how hard it is even in ideal circumstances to get everyone to vote, and how that shapes the people that end up winning elections. This article is about why campaigns undertake field efforts and how they work. It is about how campaigns battle inertia and apathy and obstacles just to make sure the people who support them actually bother to vote.

This is Field 101.

This is what 101? What's "field?"

Oh, sorry.

Did you get a phone call over the past few months? Did someone knock on your door? That's part of some campaign's field program.

Field is how campaigns refer to that direct outreach to voters. Work out in the field, as compared to television ads or sending direct mail.

Oh, you mean like lawn signs?

Ugh. No.

Some people consider lawn signs and passing out leaflets on the street and waving signs at cars part of a field campaign. Those people are wrong. Lawn signs and flyering are referred to as "visibility," getting the candidate's name out there. Visibility is demonstrably useless and annoying to voters. So why does it still happen? Because political consultants can make a good profit designing and printing those materials, and because candidates adore seeing public displays of support. (Remember how I said running for office was tedious?)

You know who helped lead the charge on dumping lawn signs? Rick Perry. If something is too stupid for Rick Perry, well… just think about that.

Visibility isn't targeted. And as you'll see, targeting voters is the heart of a field campaign.

OK. So what's "good" field?

A field program is meant to do two things: persuade likely voters to vote for your candidate and convince potential supporters to actually go out and vote.

To identify those two groups of voters, campaigns poll. They test persuasion messages (and counter-arguments to their opponents' persuasion messages) and, to a lesser extent, likelihood of voting. The results of the poll are matched demographically (and with varying sophistication) to the voter rolls, publicly-available lists of who has voted and when and where they live.

Campaigns come up with a profile of likely and possible voters based on that voting history. This year, for example, I'd call a likely voter someone who voted in November of 2004 and 2008, in the primary this year (if Republican), and in November 2010. A possible voter might be someone who only voted in 2008. An unlikely voter is someone who voted in 2004 but not 2008, or, sad as it is to say, someone newly registered to vote. (This is partly because new voters have no voting history, so it's not smart to bank on their getting to the polls.)

Once voters have been sorted into those pools -- likely to vote but not yet supportive; supportive but not yet likely to vote -- the campaign begins its field work.

For the early part of the campaign, the emphasis is on persuasion, on what are called IDs. Voters are called or reached at their doors, given a quick (poll-tested) rap, and asked if they will vote for the candidate. A "yes" respondent goes in the bank to be turned out on Election Day. An "undecided" may get another ID call later in the cycle. A "no" respondent is crossed off the list.

When polling opens, field switches over to turnout -- to GOTV, or "get out the vote." All of those "yes" respondents are now encouraged to actually cast a ballot, and are hassled -- er, encouraged -- until they do. Campaigns will also do a "blind pull," turning out voters who haven't been ID'd but who polling suggests are likely to support the candidate. For the Obama campaign, for example, it makes sense to encourage African-Americans to vote no matter what, since they support his candidacy so strongly.

This is why visibility is so stupid. If you're sent to hand out flyers at a bus stop early in the morning, you don't know if the people you're handing material to are registered voters. You don't know if they're in the right district for your campaign. And you don't know if the message in the flyer will actually persuade the person to vote for your candidate.

So what does work?

It's only been fairly recently that campaigns and outside researchers have actually looked at the most effective ways to do field outreach. The rule of thumb is that field can make a difference of two percent in the outcome of a race -- meaning that the persuasion and additional turnout can boost your final vote total by two percent. But that's not a hard and fast figure; in fact, I've never seen it researched at all.

Some of the best research on field has been done by a group at Yale University. Focused on GOTV, they provide an overview of what they've learned:

  • Personalized messages are more effective.
  • Mass email is useless for increasing turnout.
  • The content of a message can be less important than how well it is delivered.
  • In-person contacts are far more effective than phone contacts.
  • Social networks can influence people to vote.

This last point doesn't necessarily mean online social networks, though a massive experiment using Facebook to encourage turnout demonstrated that online networks can do the trick. It also means offline networks. A series of experiments (including a large one by MoveOn this week) uses public voting histories to provide voters with a list of neighbors likely to head to the polls. Research suggests that knowing that the guy next door and the family across the street voted can spur an occasional voter to actually do it. The Obama campaign recently offered another tool, allowing you to see how many people with your first name have already voted. This is probably not as effective as a network or social contact closer to you, but it works on the same principle.

(If you find the science of field campaigns interesting, and you should, check out Sasha Issenberg's book, The Victory Lab.)

How does the advent of early voting change GOTV?

Here's how Election Day used to work.

Campaigns (with enough volunteers and resources) would send "poll-watchers" into precincts to drag yes IDs and likely supporters to the polling place. Not literally drag, mind you, but everything short of it: cajole, offer a ride to, whatever it takes. Over the course of the day, campaigns (in states where it's allowed) check to see who has already voted, then go back out to focus on those who haven't yet done so. It's a time-sensitive winnowing down of a large group of possible votes into a desperate push for a handful of people as the clock ticks down.

Early voting means this whole process stretches out much longer. It's still basically the same deal -- contact the voter, encourage them to go vote, follow up if they don't -- but over weeks instead of hours. Campaigns generally love early voting because it means they can bank votes, cross them off the list, well before election day.

The presidential campaigns and associated pundits pay a lot of attention to early voting data. (See this massive dataset from George Mason University if you'd like to play along.) It can provide insight into how the campaign is going, particularly this year, when Obama's proven 2008 data can be compared to how he's doing now.

This all seems awfully slow.

It is.

A field program is by its very nature incremental, a very slow adding up of individual votes over time. Nothing reinforces the scale of modern politics, even at a local level, like participating in field. It makes obvious how overwrought claims of voter fraud would be even if fraud actually existed. Moving even a hundred votes is tedious and slow -- and makes only a minimal difference in the long run.

Is one political party better at field than the other?

Yes. Democrats are generally much better at field campaigns than Republicans, for two reasons: necessity and tradition.

You've no doubt heard that higher turnout elections generally benefit Democrats. While not a universal rule, of course, it's generally true. The reason is that the base of people who always votes will always vote. Higher turnout elections bring out additional people, who don't fit the profile of a highly likely voter.

That profile: well-off, older, white. In other words, Republican. Republicans are traditionally much more likely to vote regardless of any outside incentive, meaning that for the GOP, building a robust system to do turnout has been unnecessary. You don't need to focus on getting out the vote when the vote is already getting out on its own.

Here's what that looks like at a national level. Below is a graph showing Democratic (blue) and Republican (red) vote totals in the last nine House elections. The total turnout is that fluctuating yellow background, higher during presidential years, lower during off years. In the graph below, higher turnout doesn't necessarily help Democrats.

The above graph showed only voting in House races. The one below swaps in tallies from the presidential race during presidential election years -- for which vote totals are even higher.

Same years, different results. Look at 1996 and 2000 in particular. In presidential races, with more votes cast, Democrats did better. There are a million (more than a million, really) factors at play, but there is certainly a correlation.

Then there's tradition. For decades, labor and Democratic campaigns have worked in concert, with organized labor turning out volunteers to do the door knocking and phone calling. (I ran labor and Democratic campaigns for years in California.) Organized labor is organized and focused on organizing, making field a natural extension of what they do. In The Times, Steve Greenhouse outlines how labor is once again unrolling its field operation on behalf of President Obama.

So who's winning this year?

Each party and campaign will make a case for why it's doing better. But the answer is Obama. For the long version of why, see this article by The Atlantic's Molly Ball, which is one of the best overviews of a campaign field program I've ever read.

A subtle giveaway in that piece is here:

Instead of campaign offices, the [Republican National Committee] likes to tout voter contacts, a metric that includes the doors knocked and phone calls made by volunteers, and which cannot be independently verified. (It still counts if no one answers the door, I was told, because the canvassers leave literature behind.)

Did you see it? Yes, the data can't be independently verified. Campaigns never reveal their field data, for obvious reasons. But that parenthetical gives away how soft the field program is. Canvassers leaving literature ("lit," in the parlance) isn't voter contact. It's visibility. You don't know if your targeted voter will read that lit. You don't know if it will ever make it into the house. That doesn't count. And if visibility is mixed in with field data? Take it with a massive grain of salt.

I heard someone say the campaign "all comes down to turnout." Is that true?

Well, yeah. Pundits love to say that, and it is true -- political campaigns all come down to, you know, who actually votes. But when you hear that, you are welcome to roll your eyes. Over the course of the last weekend of the campaign, pundits run out of things to talk about since there are fewer polls and new ads and so on. So they say, "It all comes down to turnout!" Try it. It's fun.

What should I pay attention to? Can I tell from field who will win?

No. As I mentioned here, you'll want to watch trends as results start to come in. Campaigns will be scrambling to turn folks out, but that's mostly invisible to you at home and so incremental that it wouldn't tell you much anyway.

Which brings me to the most important question: How do I get campaigns to stop calling?

Here's a little secret. Campaigns know they bug you. They do. They are aware that the calls and ads and mail and everything is too much. It's aggravating. And they don't care. Because that, unfortunately, is what it takes to get people to vote: constant nagging and hassling and reminders and prompts. If you tell a campaign that you are sick of their calling and they lost your vote, they will apologize and hang up and then laugh at your naivete. Well, maybe not laugh, but not take it seriously.

There's only one way to stop that hassling, those constant calls encouraging you to vote. I'm going to tell you the secret, but you have to promise not to abuse it. You ready?

Tell them you already voted.