Want Higher Turnout? Go Walk the Neighborhood

Posted September 9, 2004 at 2:53pm

From robotic telephone calls and e-mail to door-hangers and direct mail, campaigns have a wide variety of voter-mobilization methods at their fingertips. But according to two Ivy League professors, old-fashioned shoe-leather campaigning still beats out the more technologically advanced competition.

“Face-to-face canvassing raises turnout by 7 to 12 percentage points,” Yale University professor Donald Green told a group of activists earlier this summer.

Moreover, door-to-door campaigning has a relatively low cost of about $19 per vote, said Green, who co-wrote, with fellow Yale professor Alan Gerber, a recently published book called “Get Out the Vote! How to Increase Voter Turnout.”

More expensive tactics, such as direct mail and certain types of phone banking, cost as much as $200 per vote.

In addition to cost, face-to-face canvassing offers candidates other benefits, according to Green and Gerber’s research in such states as Connecticut, California and Indiana.

For starters, door-to-door campaigning has a “spillover” effect within households. Every 10 direct contacts brings, on average, an additional six indirect contacts — that is, individuals who receive second-hand information from visits made to a spouse, parent or roommate.

“That’s one of the most exciting findings of the book,” Green said.

Green also detected evidence of “habit formation.” A single knock on the door can have a carryover effect from one election to the next in roughly one of three cases.

“A lot of GOTV focuses on persuasion,” Green explained. “If you can raise turnout in one election, you can likely raise turnout in the next election.”

That’s good news for Congressional candidates who face re-election every two years and are therefore stuck in a state of perpetual campaigning.

Not all of Green and Gerber’s research contained good news for campaigns. The academics found that several frequently used, and often expensive, get-out-the-vote tactics may be far less effective than once thought.

“Commercial phone banks have relatively little effect,” said Green. He said that past research has never found the percentage boost from such efforts to be statistically significant.

Green said he also studied the use of robotic phone calls — including a study of turnout in the race by one Latino Congresswoman — and discovered a “negligible, if any, effect.”

One experiment involving leaflets containing information about the voter’s polling place raised turnout by about 1 percentage point. Another study of “nonpartisan” direct mail resulted in about a half percentage point boost, he said.

While leafleting may be “easier, faster and considerably less demanding than door-to-door canvassing,” the book reports, finding a leaflet at one’s door is a less memorable experience than having a face-to-face conversation with a canvasser.

That said, the book warns that not just any door-to-door canvassing campaign will do.

GOTV canvassing isn’t always effective in densely populated areas because one campaign often duplicates the efforts of others. Even more surprising, Green said, “it turns out that the more you do, the less” effective the effort may be.

The book, in fact, details the failure of a nonpartisan canvassing campaign in South Bend, Ind., prior to the 2002 elections. Their conclusion: The voters may have been overwhelmed.

“Battling over a contested Congressional seat, both parties apparently canvassed the same turf chosen by the nonpartisan campaign, which may have caused voters to become saturated with GOTV appeals,” Green and Gerber wrote.

The authors also noted that door-to-door canvassing for voter mobilization can be improved if there’s a “match” between the canvasser’s ethnic profile and that of the neighborhood.

One get-out-the-vote campaign in Raleigh, N.C., that utilized both Caucasian and black canvassers in a predominately white area encountered significant difficulties, the book noted.

“According to canvassers’ reports, some white residents refused to open their door to black canvassers,” the book stated, and two black canvassers were “accosted by white residents and told to leave the neighborhood.”

Gerber and Green’s research in Dos Palos, Calf., revealed that when a team of Latino Democrats were randomly assigned to talk to both Anglo and Latino voters, they had much better luck among Latinos.

Some consultants offered a slightly different take on the issue.

Jim Jennings, a longtime Democratic campaign strategist who specializes in get-out-the-vote efforts, said that while door-to-door canvassing is the ideal method of contacting voters, it can be extraordinarily difficult to mount a successful canvassing outfit.

“There are many, many neighborhood where door-to-door can’t be done without seriously offending the people who live there,” he said. Moreover, in a presidential election, it is be next to impossible to amass “enough bodies” to ring every doorbell.

In his experience, Jennings said, the “next best thing up from a human being at the door” is a piece of direct mail, particularly when followed up by a phone call.