Does Early Voting Really Improve Voter Turnout The View from the Academy

1
Does Early Voting Really Improve Voter
Turnout?The View from the Academy

• Paul Gronke, Director
• Eva Galanes-Rosenbaum, Research Director
• Early Voting Information Center
• July 16, 2007
• Prepared for the National Association of
  Secretary of States Summer Meeting, Portland OR

2
Who am I?

• Professor of Political Science at Reed College
  and Director of the Early Voting Information
  Center
• Primary fields of research are voting behavior,
  public opinion, survey administration and
  statistics
• Recent projects
• Carter/Baker Commission report on Voting by Mail
• 2006 Carnegie Corporation project on 2004
  Election Day Survey
• 2007 EAC Contract on 2006 Election Administration
  Survey
• Where you can find out more about me
• Email earlyvoting_at_reed.edu
• Web http//earlyvoting.net, http//www.reed.edu/
  gronke
• Blog http//electionupdates.caltech.edu

3
Agenda for today

• Theories of voting and the early voter
• Voting as A Cost/Benefit decision Because EV
  decreases cost, turnout should increase
• Voting as Consumption Because voting is a
  marginal decision, EV should have small effects
  (positive and negative) on turnout
• Voting and the Social Context If states make it
  easier for parties, candidates, and other
  organizations to identify early voters, turnout
  should increase
• Empirical studies of turnout and early voting
• How scholars and policymakers can work together

4
State of the World Early Voting is coming and
you cant stop it
6 states cast ballots before Iowa Caucus
12 states cast ballots before New Hampshire
Major share of delegates at stake in early voting
Thanks to
Source Fitzgerald 2005
5
Voting as a Cost-Benefit Decision

• Why should anyone vote at all?
• Scholars often conceptualize voting as a
  cost-benefit calculation
• Vote if C lt pB
• The probability that a citizens vote will make a
  difference (p) multiplied by the perceived
  benefit from voting (pB) must be greater than the
  cost of voting (C) in order for a citizen to
  vote.
• Prediction rational citizens should not vote and
  in equilibrium, turnout should be zero

6
How to increase turnout if you believe the
Cost-Benefit Model

• The p (probability that one vote will make a
  difference) is virtually nil thus
• Efforts to increase turnout work on B or C
• B most GOTV efforts for candidates and policy
  campaigns work here, trying to make potential
  voters feel that the benefit gained from voting
  is large (or that the cost of not turning out
  will be severe)
• C a few campaigns, and most civil servants
  operate here, trying to lower the costs of voting
  by making polls easier to access, making
  information accessible, etc.
• Early Voting may increase turnout by lowering the
  costs of voting.

7
Voting as consumption

• Many political scientists modify the classic
  cost-benefit model of turnout
• Vote if C lt pB D
• D has been variously described as civic duty or
  value of democracy.
• Consumption because you enjoy the benefit
  whether or not your candidate wins. You
  consume voting.
• Prediction turnout should equilibrate at some
  level greater than zero, and vary according to
  the consumptive value of the election.

8
How to increase turnout if you believe the
consumption model

• Manipulate D
• Encourage a sense of civic duty
• Describe voting as a way to express your voice
  (Rock the Vote), not to win or lose for policy
  purposes
• Structure voting so as to encourage community
  activity and civic rituals
• Early voting
• Should have no impact, because small changes on
  the cost side wont translate into big changes on
  the turnout side
• Should have a negative impact because it destroys
  old civic rituals
• Should have a positive impact because it creates
  a new civic ritual (voting at the kitchen
  table).
• Regardless, any effects are likely to be small.

9
Voting and the Social Context

• Criticism of homo economicus. What about homo
  societushumans embedded in a social context?
• Institutions play a role by getting people out to
  vote
• Institutions play a role by teaching skills and
  providing resources needed to help navigate
  political waters
• Evidence
• Mobilization can overcome barriers to voting
• Citizens with greater social connections vote
  more frequently
• Church attendance
• Members of civic organizations
• Volunteerism

10
How to increase turnout if you believe in
mobilization

• Encourage the health of American civic life
• Fund political parties and not just candidate
  organizations
• Encourage volunteerism and civic engagement no
  more bowling alone
• Early voting
• May have a positive impact if states and
  jurisdictions make permanent absentee lists and
  early voting returnseasily and cheaply available
  (statewide?)
• Allows civic organizations to target mobilization
  efforts
• Promotes cost efficiency among candidates and
  party organizations
• May level the campaign playing field and lower
  campaign spending

11
VBM and Turnout Empirical Evidence

• Academic verdict on VBMs effect
• Small increase in turnout
• EVIC estimates are that voting by mail has
  increased turnout in Presidential elections
  approximately 5
• One California estimated a decrease of 2.6
• VBM does not expand the electorate
• Tends to draw in regular voters, does not attract
  new voters
• What we dont know yet
• Changing landscape of voting by mail
• Concerns about novelty effect and other
  special circumstances
• Effects likely vary in higher and lower profile
  contests

12
Other Modes and Turnout

• Liberalized absentee voting
• EVIC compared primary types of non-precinct place
  voting since 1960, in Federal elections, and
  found no impact (other than VBM)
• Other studies have found consistently positive
  but small (less than 3) and often statistically
  insignificant
• In Person Early Voting and Voting Centers
• No measurable impact on turnout
• Early voting allows parties to harvest core
  voters
• What we dont know yet
• Based primarily on studies in Texas and Colorado
• Widespread party mobilization efforts directed at
  early voters

13
Who votes early? The Early Electorate

• Resource Rich voters (better educated, higher
  income)
• Committed voters (voters with strong ideological
  leanings and partisan ties, older voters?)
• Voters who are inconvenienced by traditional
  precinct place voting (e.g. voters facing longer
  commutes, older voters?)
• Early voting does not expand the electorate and
  reinforces existing biases in American politics
• Early voting methods increase voter turnout on
  average (across many different types of
  elections), primarily by making it easier for
  current voters to continue to participate, rather
  than by mobilizing nonvoters into the electorate.
  
• Very consistent with the theoretical
  expectations

14
Looking to the future

• Electoral reforms that make voting more
  convenient but may exacerbate pre-existing biases
• Retention is easy to fix. Likely voters1 who
  dont participate already are those who respond
  best to lower physical coststheyre already
  engaged, wealthy, educated, or habitual voters
• Expanding the electorate is hard.
• Turning unlikely voters into likely voters
  involves much more than administrative changes
• 1. Likely voters are those who share
  characteristics with habitual voters older,
  wealthier, politically and socially engaged,
  educated, etc. The expectation that a member of
  these groups will turn out to vote is higher than
  for others.

15
How EV research can help you

• Study EV as a way to improve election
  administration
• Give poll workers extra time to learn in a
  hands-on environment
• Allow many problems to be caught and dealt with
  before the huge crowds of Election Day
• Give voters the option of returning later if
  something isnt working or if access is a
  problem
• Study EV as a way to increase turnout and expand
  the electorate
• Experimental studies of different methods of
  voter outreach and voter assistance
• Different methods for different citizens
• Study EV across jurisdictions to establish best
  practices

16
How you can help research

• Tracking the number of early votes in an election
  is a source of considerable frustration
• States such as Texas, Oregon, Florida make data
  readily available, in real-time and permanently
• Other states dont make these numbers available
  easily
• Often cant produce the data when directly
  solicited
• Many dont track EV at the state level (must go
  to each county)
• Inconsistent definitions
• Quick collection and analysis of national EV
  rates is currently impossible
• Take full advantage of statewide voter
  registration systems and link to voter history
  files