Nonpartisan
GOTV
Canvassing and Voter Turnout
1998 Connecticut Congressional and State Elections
What we
did
In
New Haven, Connecticut, a city with a population of 100,000, we drew a
sample of nearly 30,000 registered voters for a study of GOTV mailings,
phone calls and door-to-door canvassing. We tested the effectiveness of
canvassing by hiring graduate student canvassers, who, when possible,
were matched to the racial and ethnic composition of the neighborhoods
they walked. During each Saturday and Sunday for four weeks before the
election, canvassers were sent to contact randomly selected registered
voters. They conveyed one of three different GOTV messages: a civic duty
version; a close election version; and a neighborhood solidarity version.
What we found
After
accounting for the higher probability of voting among those who are easier
to contact, we found that door-to-door canvassing raises the probability
of turnout by 9.8 percentage points. Differences in the content of the
GOTV message did not yield statistically significant differences in likelihood
of voting.
2000 Multi-city Presidential Year Elections
What we
did
Young people, ages 18-30, were contacted other young people in a door-to-door
canvassing campaign undertaken by the Youth
Vote Coalition. The canvassing campaign took place in Eugene, Oregon,
Ann Arbor, Michigan, and Boulder, Colorado prior to the 2000 election.
Volunteers talked with potential voters using an informal, chatty style.
Our total sample for this study included over 25,000 registered voters.
What we
found
Face-to-face contact raised turnout in the treatment group by as much
as 8.5 percentage points. In addition, we found that face-to-face contact
with this group produced spillover effects. That is, the intervention
even had a positive impact on voter turnout among other household members
of the treatment group, raising their turnout by 2.7 percentage-points.
Thus, for every 100 treatment subjects assigned to a canvassing campaign,
4.3 additional votes were mobilized through intended contact and another
4.1 votes were mobilized through incidental contact, or spillover.
2001 Multi-city Local Elections
What we
did
This study was also conducted in conjunction with the Youth Vote Coalition.
For the 2001 local off-year elections the Youth Vote Coalition (drawing
on volunteers from local PIRG
chapters and ACORN)
sought to mobilize young voters in Bridgeport, Connecticut, Columbus,
Ohio, Detroit, Minneapolis, and Raleigh.
What we found
All
canvassing efforts were successful with the exception of that in Raleigh
- which had several unusual problems and complicating circumstances. In
the other five sites, however, actual contact with a canvasser increased
turnout among all age groups by an average of 10.9 percentage-points.
And among voters 25 and younger, face-to-face canvassing had an even stronger
effect, raising turnout rates by 12.3 percentage-points. As seen in 2000,
in addition to the intended effects on those individuals targeted by canvassers,
the researchers found that voting also increased for other members of
the individuals' households. Across all of the sites, excluding Raleigh,
the estimated spillover effect was an impressive 5.7 percentage points.
We found that canvassing appeals that were coupled with issue advocacy were slightly more effective in getting voters to the polls than straight GOTV messages. However, we found that providing voters with voter guides had no discernable influence on turnout.
The Bottom
Line
Good
old fashioned door-to-door canvassing seems to be the clear winner when
compared with nonpartisan phone calls and direct
mailings. If $1.50 is the cost per personal contact (10 contacts per
hour at $15 per hour), then we estimate approximately one more voter is
"delivered" for each $8 spent. In contrast, at fifty cents per
mailer, sending three mailings to each household (containing an average
of 1.5 voters) nets roughly one additional voter for each $40 spent. Our
bottom line advice is, think very carefully about how you allocate your
GOTV resources - phone banks and mailings may seem cheaper upon first
glance, but our research shows that intensive canvassing is actually far
more cost-effective.
Partisan
GOTV
Canvassing and Voter Turnout
Please check in later. These studies have not been completed.
For more
information on our studies, go to www.yale.edu/isps/publications/voter.html
or contact Beth Weinberger at beth.weinberger@yale.edu.
GET
OUT THE VOTE
URL: http://www.yale.edu/vote/canvassing.html
Certifying authority: Donald P. Green
Last reviewed: 11.17.05